www.nature.

com/scientificreports

OPEN Improved SSD network for fast

concealed object detection
and recognition in passive
terahertz security images
Lu Cheng1,2,3, Yicai Ji1,2,3*, Chao Li1,2,3, Xiaojun Liu1,2 & Guangyou Fang1,2,3
With the strengthening of global anti-terrorist measures, it is increasingly important to conduct
security checks in public places to detect concealed objects carried by the human body. Research
in recent years has shown that deep learning is helpful for detecting concealed objects in passive
terahertz images. However, previous studies have failed to achieve superior accuracy and performance
for real-time labeling. Our research aims to propose a novel method for accurate and real-time
detection of concealed objects in terahertz images. To reach this goal we trained and tested a
promising detector based on deep residual networks using human image data collected by passive
terahertz devices. Specifically, we replaced the backbone network of the SSD (Single Shot MultiBox
Detector) algorithm with a more representative residual network to reduce the difficulty of network
training. Aiming at the problems of repeated detection and missed detection of small targets, a
feature fusion-based terahertz image target detection algorithm was proposed. Furthermore,
we introduced a hybrid attention mechanism in SSD to improve the algorithm’s ability to acquire
object details and location information. Finally, the Focal Loss function was introduced to improve
the robustness of the model. Experimental results show that the accuracy of the SSD algorithm is
improved from 95.04 to 99.92%. Compared with other current mainstream models, such as Faster
RCNN, YOLO, and RetinaNet, the proposed method can maintain high detection accuracy at a faster
speed. This proposed method based on SSD achieves a mean average precision of 99.92%, an F1
score of 0.98, and a prediction speed of 17 FPS on the validation subset. This proposed method based
on SSD-ResNet-50 can provide a technical reference for the application and development of deep
learning technology in terahertz smart security systems. In the future, it can be widely used in some
public scenarios with real-time security inspection requirements.

In recent years, the global counter-terrorism situation has become increasingly severe. Security inspections in
public places have gradually attracted widespread attention from various countries. In crowded public places,
in addition to the security checks on the packages people carry, it is also important to conduct security checks
on the human body to detect dangerous hidden objects. However, the existing security inspection systems have
some deficiencies more or less. In this case, the terahertz security inspection system using electromagnetic waves
as the detection method is playing an increasingly important role.
Terahertz waves refer to electromagnetic waves with frequencies between 0.1 and 10 THz, which have similar
characteristics to microwaves and ­infrared1. Terahertz imaging can detect not only metallic objects, but also
non-metallic contraband such as explosives, ceramic knives, drugs, and glass k­ nives2,3. Terahertz waves can
penetrate insulating materials such as clothing, plastics, and ­ceramics4–6, so they can be used to detect objects
hidden under human clothing. Compared with microwaves, the terahertz waves have shorter wavelengths and
higher imaging accuracy. Terahertz waves have much longer wavelengths compared to X-rays and visible light.
Therefore, the resolution and signal-to-noise ratio of human security imaging based on terahertz systems are
quite different from X-ray imaging and visible light imaging. Such characteristics make it difficult for terahertz

1
Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, China. 2Key
Laboratory of Electromagnetic Radiation and Sensing Technology, Chinese Academy of Sciences, Beijing 100190,
China. 3School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of
Sciences, Beijing 100049, China. *email: [email protected]

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 1

Vol.:(0123456789)
 www.nature.com/scientificreports/

imaging to automatically detect hazardous objects using existing object detection algorithms. Terahertz imaging
systems are divided into active and passive working modes. In the application of detecting concealed objects in
the human body, passive terahertz imaging systems occupy a major ­position7,8. Passive terahertz imaging systems
do not emit electromagnetic waves and will not cause harm to the human body. The current speed of passive
terahertz imaging is as high as 10 frames per second, which puts forward higher requirements for the speed and
accuracy of detection. At present, most of the images acquired by the terahertz systems are checked manually one
by one, which not only consumes a lot of manpower, but also may lead to missed inspections when people mark
images for a long time. As a result, the advancement of automatic identification technology for passive terahertz
images is critical for the widespread adoption of intelligent security inspection scenarios.
It is well known that the security screening of terahertz imaging systems is designed to detect and identify
the threat of dangerous goods carried by the human body. More and more terahertz passive imaging systems
have been studied in recent years, but theories and methods for identifying or locating dangerous objects from
terahertz images are still in their infancy. Compared with optical and active terahertz imaging, passive terahertz
images have a low signal-to-noise ratio and are easily affected by background variations. These characteristics
make the use of passive terahertz imaging systems for security inspections a huge challenge. Traditional object
detection algorithms of passive terahertz images mainly include segmentation-based detection algorithms and
feature matching-based detection algorithms. The segmentation-based detection algorithm uses the grayscale
information in the passive terahertz image to divide the image into two regions, the target and the background,
and then detects the target. At present, the commonly used segmentation algorithms include maximum entropy
segmentation ­method9, Otsu ­method10, region growing method, etc. The target segmentation-based algorithm is
fast and easy to implement, but it can’t detect complex targets. The feature matching-based detection algorithms
construct a target feature descriptor to match known prior feature information. Common features include Haar
­features11, gradient h­ istograms12 (HOG), dense SIFT f­ eatures13, etc. Feature-based detection methods rely on
manual feature extraction. However, since artificially constructed features are usually limited to specific object
types, satisfactory object detection results cannot be achieved. In recent years, with the development of deep
learning, the application of deep convolutional neural networks in optical image target detection has achieved
good results, and its detection accuracy is far superior to traditional m­ ethods14,15. Low-resolution passive tera-
hertz images contain very little information, making it very difficult to extract features from images manually.
The CNN network can automatically extract features from passive terahertz images. Although large-scale CNN
networks have good detection and classification effects, they cannot meet real-time performance due to large
parameters and complex structures. Extensive research shows that there is still space for improvement in the
accuracy and speed of current CNN-based terahertz image target detection a­ lgorithms16–18. Hence, there is an
urgent need to propose a method with sufficient accuracy and robustness for object recognition in passive tera-
hertz images. Nevertheless, to the best of the author’s knowledge, there are only a handful of studies in the area.
There are two main problems in the direct use of deep learning algorithms in the field of passive terahertz
image object recognition, one is that the accuracy of object detection is not high, and the other is that the speed of
object detection is not fast enough. In response to these two issues, we chose the single-stage detection algorithm
SSD, which has relatively high accuracy and speed in the natural light object detection algorithm, as the basic
model. This is the first application of the SSD algorithm in the field of passive terahertz image object detection.
We investigate the potential of the residual network ResNet-50 in terms of accuracy and real-time performance
in detecting concealed objects in passive terahertz images. The method described in this paper aims at select-
ing the most efficient CNN architecture and further exploring the ways of its modification and optimization to
ensure superior real-time classification potential for detecting concealed objects. The significance and originality
of this research lie in the first exploration of the application of the SSD model in the field of passive terahertz
image detection and several important improvements are proposed for image characteristics. The ultimate goal
of this research is to develop a fast and accurate detection algorithm for passive terahertz images and promote
its practical application in intelligent security inspection scenarios. We propose the following improvements
to further improve the accuracy and speed of the algorithm in detecting concealed objects. To summarize, our
main contributions are as follows:

• Aiming at the network degradation problem of the VGGNet n ­ etwork19 in the SSD algorithm, the improved
algorithm uses the ResNet-50 n ­ etwork20 as the feature extraction network.
• For the problem of poor detection performance of small objects, we extract multi-level features to form a
feature pyramid ­network21, which is used to detect objects of different scales. Through upsampling, deep
features and shallow features are fused to construct feature representations with rich semantic information,
making the fused features more descriptive.
• The proposed algorithm introduces the spatial-channel attention mechanism into the SSD network to enhance
the semantic information of high-level feature maps. Therefore, the ability of the algorithm to obtain the
details and position information of the object can be improved, thereby reducing the missed detection rate
and false detection rate, and improving the detection accuracy of small objects.
• This paper introduces the Focal L ­ oss22 to improve the imbalance of positive and negative samples and hard
and easy samples. By increasing the weight of the hard samples in the loss function, the robustness of the
proposed algorithm can be improved.

Compared with the unoptimized SSD algorithm, this method can improve the detection accuracy and at
the same time meet the detection speed requirements in intelligent security inspection scenarios. Specifically,
our proposed algorithm achieves a mean average precision of 99.92%, and the detection speed of the algorithm
can reach 17 FPS, which can meet the real-time requirements. This is an excellent outcome that could not have

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 2

Vol:.(1234567890)
www.nature.com/scientificreports/

Figure 1.  The network structure of SSD.

been obtained with earlier methods. The remainder of the paper is organized as follows: The related works on
terahertz image object detection and the basic introduction to SSD algorithm are introduced in “Related work”.
In the following section, the proposed improved SSD algorithm is presented in detail. Section “Experimental
results and discussion” shows and analyzes the experimental results of our method after its deployment. The
conclusion of the paper is formed in the last section.

Related work
This section introduces the related works, including the introduction of common object detection methods based
on CNN models, the basic introduction of the SSD model, and related work on terahertz image object detection.

Object detection and recognition based on CNN models. In recent years, deep learning has made
breakthroughs in object visual detection. At present, object detection algorithms based on deep learning models
are mainly divided into two categories: one is various two-stage algorithms based on R-CNN23, including Fast
R-CNN24, Faster R-CNN25, Mask R-CNN26, ­RFCN27, etc. These two-stage algorithms greatly improve the detec-
tion accuracy, but the two-stage detection results in a slow detection speed, which cannot meet the real-time
requirements. The other category is one-stage detection algorithms represented by Y ­ OLO28 (You Only Look
Once) and ­SSD29. The YOLO algorithm uses the idea of regression to greatly improve the detection speed, but
its recognition accuracy of object location is poor and the recall rate is low. The SSD algorithm combines the
regression method in the YOLO algorithm, not only follows the anchor mechanism in Faster R-CNN but also
extracts multi-scale feature maps for classification and prediction, which greatly improves the precision and
speed of object detection.

Detailed introduction of SSD algorithm. The SSD algorithm is an object detection algorithm proposed
by Wei Liu et al.29, which combines the anchor mechanism in Faster R-CNN and the regression idea in YOLO.
The SSD algorithm not only has a clear speed advantage over Faster R-CNN, but also has a clear accuracy
advantage over YOLO. It is based on the VGGNet-16 backbone and fuses feature maps of different scales for
feature extraction. Its network structure is divided into two parts: one is the feature extraction network, and the
other is the classification and regression layer. Specifically, the feature extraction network can extract the feature
information of the object in the image, and at the same time can improve the ability of the network to perceive
the object. The classification and regression layers are capable of classifying and regressing each candidate box to
detect objects in the terahertz images. The structure of the original SSD network is shown in Fig. 1.
As we all know, the SSD algorithm no longer uses fully connected layers, which improves computational effi-
ciency. Moreover, the SSD algorithm uniformly modifies the size after the input image, and there is no mandatory
requirement for the size of the input image. Besides, SSD directly uses convolution to calculate candidate boxes
and predict classification in one step, which can simplify the process. However, the SSD algorithm also has some
disadvantages. To begin with, the maximum size of feature maps used for prediction has a resolution of 3838. If
the size of small objects in the input image is small, the detailed information contained in the lower layers can
easily be lost after the pooling layer. Besides, as the feature map size of additional feature extraction networks
continues to decrease, the feature extraction and representation capabilities of shallow feature layers are limited.
While the feature output layer of deep features is mainly responsible for large objects, resulting in poor detection
performance of SSD for small objects. Finally, the convolutional layer of SSD algorithm feature extraction can-
not take into account the features of adjacent feature layers of different scales. This results in insufficient feature
extraction capability for small object features in complex environments during object feature extraction. All in
all, the detection performance of the original SSD algorithm for small objects in complex environments needs
to be improved. To meet the need of detecting hidden objects in passive terahertz images, some improvements
to its network structure are also required.

Hidden object detection in terahertz security images. In recent years, passive terahertz imaging
technology has shown significant development prospects in the field of security inspection. It is of great sig-
nificance to detect dangerous objects hidden in the human body in passive terahertz security images. At pre-
sent, some scholars have carried out research on passive terahertz human body security image object detection
­algorithms30,31. According to the characteristics of terahertz images, Zhang et al. adopted the Otsu threshold
segmentation algorithm and applied the contour tracking method to extract contours, which can realize object

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 3

Vol.:(0123456789)
 www.nature.com/scientificreports/

detection in human terahertz ­images32. Aiming at the difficulty of detecting edge objects in human terahertz
images, Jiang et al. proposed an automatic recognition algorithm for human edge ­objects33. Santiago Lopez
Tapia et al. proposed a method combining image processing and statistical machine learning techniques to solve
the problem of object localization detection in terahertz ­images34. Niu et al. proposed a terahertz human image
processing and recognition method based on the principle of saliency and sparse coding, which can realize auto-
matic recognition of hidden objects in the human ­body35.
In these methods, the objects were first separated from the image and then classified and recognized in a
complex and slow process. Moreover, the above traditional object detection methods have certain defects and
poor generalization ability. The performance of these algorithms is often affected by the complexity of the image
background. The simpler the image background, the higher the object detection efficiency and the better the
detection performance. On the contrary, once the image background becomes complex, the efficiency and per-
formance of object detection will decrease accordingly.
The convolutional neural network ­CNN36 (convolutional neural network) based on deep learning technology
can solve the defects of traditional methods well. CNN can not only complete feature extraction, but also has good
robustness and strong feature expression ability. These properties enable it to accurately localize detection objects
in both simple and complex environments. Liu Chen et al. introduced Focal L ­ oss22 on the basis of Faster R-CNN
and detected concealed objects in active millimeter-wave images without using transfer l­earning16. Hong Xiao
et al. proposed the R-PCNN algorithm. The algorithm adds traditional image preprocessing methods to the front
end, which improves the speed and accuracy of object detection and r­ ecognition17. Xi Yang et al. exploited the
spatiotemporal information of terahertz image sequences and used CNN to achieve automatic object detection
and ­recognition18. Extensive studies show that there is still space for improvement in the accuracy and speed of
current CNN-based terahertz image object detection algorithms.

Methods
According to the characteristics of terahertz images, we propose a novel method for object detection in terahertz
human security images. This method enables accurate real-time detection of concealed objects in terahertz
images. The improved SSD framework is shown in Fig. 2. In general, the improved SSD object detection algorithm
is divided into the following four parts. Firstly, the basic network VGGNet-16 is replaced with a deep convolu-
tional network ResNet-50 to enhance the feature extraction ability. Additional feature layers are introduced into
the basic network to further enhance the feature expression capability of the object detection layer. Afterward,
three feature maps of different scales are fused in the feature extraction layer to enhance the semantic information
correlation between the front and rear scale maps. In the next parts, a hybrid attention mechanism is introduced
into SSD to enhance the semantic information of high-level feature maps. This method can improve the algo-
rithm’s ability to obtain object details and position information, thereby reducing the missed detection rate and
false detection rate. Finally, the Focal Loss function is introduced to improve the robustness of the algorithm by
increasing the weight of negative samples and hard samples in the loss function.
The overall flowchart of our proposed algorithm is shown in Fig. 3. During the training phase, the objects in
the input image are marked to obtain the location information and category information of the real target. Then
the model is trained to generate the final improved SSD object detection model. During the testing phase, each
test image generates N boxes that may contain objects. Ground truth offsets and class scores are then computed
using the model generated during the training phase, resulting in N classification results per image. Finally, the
non-maximum suppression algorithm is used to output the final result.

Ethical statement. This study conforms to the ethical guidelines of the Declaration of Helsinki revised in
2013. The study was approved by the ethics committee of the Aerospace Information Research Institute, Chinese
Academy of Sciences. All experiments were performed by relevant guidelines and regulations. We confirmed
that informed consent had been obtained from all subjects. All images were deidentified before inclusion in this
study.

Improvement of feature extraction network. The VGGNet-16 network is a simple stack of traditional
CNN structures with many parameters. As the number of network layers deepens, there are not only the phe-
nomenon of gradient disappearance and gradient explosion, but also the problem of network degradation. In
response to the problems existing in the traditional CNN structure, the team of Kaiming He of MSRA proposed
a convolutional neural network called Deep Residual Network(ResNet)20, which successfully solved the above
two problems by introducing a batch normalization (BN) layer and residual block. The BN layer normalizes the
feature maps of the middle layer to speed up the network convergence and improve the accuracy. The residual
block fits the residual map by adding a skip connection layer between the network layers. It can directly transfer
the optimal weight of the front layer to the back layer, so as to achieve the effect of eliminating network degrada-
tion. The comparison between the ordinary structure and the residual structure is shown in Fig. 4. The ResNet
network adopts the jump structure in Fig. 4b as the basic network structure, also known as Bottleneck. The jump
structure enables the ResNet network to have deeper layers and relatively better network performance than
ordinary networks. As shown in Fig. 4, H(x) is called the desired map for stacking several network layers, x is
the entry of the current stacking block, and the use of the ReLU activation function shortens the learning cycle.
Assuming that n nonlinear layers can be approximately expressed as a residual function, then the stacked net-
work layers are fitted to another map F(x) = H(x) − x , and the final base map becomes H(x) = F(x) + x . The
idea of residual structure can effectively solve the problem of network performance degradation and gradient,
and also improve network performance. By constructing residual learning, the residual network can approxi-
mate a better-desired mapping by approximating the coefficients of multiple nonlinear connections to zero.

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 4

Vol:.(1234567890)
www.nature.com/scientificreports/

Figure 2.  Improved SSD network architecture.

The detection of concealed objects in passive terahertz images is affected by complex backgrounds and similar
interference. Research has proved that deepening the network structure is effective for improving feature detec-
tion. However, VGGNet-16 is a relatively shallow network, which is insufficient to extract features of hidden
objects. Therefore, we replace the feature extraction network with ResNet-50, a network with deeper and richer
semantic information. ResNet-50 has a total of 16 Bottlenecks, each of which contains 3 convolutional layers.
Together with the input layer and the final fully connected layer, a 50-layer residual network is formed. The com-
parison of network parameters and complexity between VGGNet-16 and ResNet-50 is shown in Table 1. It can be
seen from Table 1 that the parameters and floating-point operations of the ResNet-50 network are much lower
than those of the VGGNet-16, which proves that ResNet-50 is lighter and has the conditions for faster speed.
Compared to AexNet, GoogleNet, and VGGNet-16, ResNet-50 is known for its ability and performance to
achieve lower error rates and higher accuracy output than other a­ rchitectures37. The depth of each convolu-
tional layer of ResNet-50 is higher than that of VGGNet-16, which enhances the network learning ability and
further improves the feature extraction ability of concealed objects in images. ResNet-50 introduces a residual
module in the convolutional layer, which solves the problem of training degradation caused by the deepening
of the network. The network has high convolution calculation efficiency and reduces the redundancy of the
algorithm. Taking into account factors such as accuracy and detection speed, the novel CNN network based
on the ResNet-50 network is more accurate than VGGNet-16. By using ResNet-50 as the backbone network,
the data features can be better extracted for classification and loss computation. The improved algorithm uses
ResNet-50 as the extraction network, removes the fully connected layer in ResNet-50, adds several additional
convolutional layers to extract features, and obtains feature maps of different scales. In this way, the limitation
of the fully connected layer on the size of the input image can be solved. The network can obtain feature maps of
different sizes to construct the detection part of the feature maps of different sizes in the network. The size of the
feature map of the feature extraction layer of the improved algorithm is 3838, 1919, 1010, 55, 33 and 11, and the
dimensions are 128, 256, 512, 256, 128 and 64 respectively. The shallow convolution layers extract relatively small
features, while deep convolution layers can extract richer information. Therefore, shallow convolutional feature
maps and deep convolutional feature maps are used to detect small and large objects, respectively. Replacing the

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 5

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 3.  Flowchart of the algorithm.

Figure 4.  Comparison diagram of traditional structure and residual structure. (a) The traditional structure. (b)
The residual structure.

original front-end network VGGNet-16 with the ResNet-50 network can improve the accuracy on the basis of
the original SSD algorithm, but there are still problems such as poor real-time performance, false detection, and
repeated detection of small objects. Therefore, feature fusion techniques are used to improve the SSD algorithm.

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 6

Vol:.(1234567890)
www.nature.com/scientificreports/

Network Parameter quantity Number of floating-point operations

VGGNet-16 138 15.5
ResNet-50 25.5 3.8

Table 1.  Comparison of network complexity between VGGNet-16 and ResNet-50.

Figure 5.  Overview of the proposed feature fusion strategy.

Feature fusion module. The SSD algorithm takes full advantage of multiple convolutional layers to detect
objects. It has better robustness to the scale changes of objects, but the disadvantage is that it is easy to miss small
objects. The network structure of the SSD algorithm is shown in Fig. 1. Each convolutional layer corresponds
to the scale of an object, thus ignoring the relationship between layers. The low-level feature layer Conv_3 has a
resolution size of 3838 and contains a lot of edges and non-target information. Conv_4 and Conv_5 have higher
resolution and basic contour information and thus have richer and more detailed feature representation. As
the number and depth of network layers increase, the scale of convolutional layers gradually decreases, and the
semantic information becomes more and more abundant. However, the underlying conv4_3 does not utilize
high-level semantic information, resulting in poor detection of small objects.
The improved algorithm fuses adjacent low-level high-resolution feature maps with clear and detailed features
and high-level low-resolution feature maps with rich semantic information. Feature fusion can enhance the abil-
ity of the low-level feature layer to fully express the detailed features of small objects, and improve the detection
effect of the SSD algorithm on small objects. The feature fusion structure proposed in this paper is shown in
Fig. 5. The improved SSD model performs upsampling operations on Conv_4 and Conv_5 by bilinear interpo-
lation, and performs feature fusion with Conv_3. Better upsampling results can be obtained by using bilinear
interpolation operations instead of adjacent interpolation operations. We use the concatenation method to fuse
the high-semantic feature maps of Conv_4 and Conv_5, and use them to supplement the feature information of
the convolutional feature maps of the Conv_3. Moreover, we still use the scale of Conv_3 as the scale of the new
prediction stage. This way of passing down the fusion information can further fuse high-semantic information
in deep layers and detailed information in shallow layers.
As we all know, the update of the parameters of the previous convolutional layer will cause the data distribu-
tion of the later input layer to change, resulting in a large difference in the data distribution of the convolutional
feature layer. Therefore, there will be large differences between feature maps, and feature connection operations
cannot be performed directly. When the network layer changes slightly, the features are accumulated and ampli-
fied through the fusion layer, resulting in slow convergence of the algorithm. Therefore, a BN layer is added after
each feature layer for normalization. BN layers can reduce the effects of a slow training process due to increased

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 7

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 6.  The overview of CBAM attention mechanism module.

Figure 7.  Diagram of channel attention module.

model complexity. The improved SSD algorithm adopts the method of multi-level fusion of different scales,
combined with the idea of ­FPN21, and transfers the feature information of feature maps of different scales from
top to bottom. Feature fusion can provide feature representations with rich semantic information, improving
the descriptiveness of fused features.

Attention mechanism CBAM. The feature maps extracted by the feature extraction network not only
have object features but also have similar interference features. The concealed objects on the human terahertz
image and other similar objects are given the same importance on the feature maps, which is not conducive
to the detection of hidden objects in complex backgrounds. Therefore, to improve the object recognition abil-
ity of feature maps for specific regions and specific channels, and reduce the negative interference of complex
backgrounds and similar objects, the attention mechanism is applied in both channel and space dimensions.
Mechanism of Attention C ­ BAM38(Convolutional Block Attention Module) is a little universal module that saves
computational resources and parameters. For a given feature map, attention weights are sequentially derived
along both spatial and channel dimensions, and then the features are adaptively adjusted by multiplying with the
original feature map. The implementation of the attention mechanism includes two modules: channel attention
and spatial attention. The structure of the CBAM attention mechanism module is shown in Fig. 6. Usually, the
two modules of channel attention and spatial attention are combined in turn, and better results can be achieved
by placing the channel attention in the front. The improved algorithm connects the CBAM dual attention mech-
anism after each output feature map to improve the network’s attention to concealed objects.
Given an intermediate feature map F ∈ RC×H×W as input, CBAM sequentially infers a 1D channel atten-
tion map Mr ∈ RC×1×1 and a 2D spatial attention map Ms ∈ R1×H×W as shown in Fig. 6. The overall attention
progression can be summarized as:
F ′ = Mc (F) ⊗ F (1)

F ′′ = Ms F ′ ⊗ F ′ (2)
 

where ⊗ represents for element-wise multiplication. During multiplication, the attention values are broadcast
accordingly: channel attention values are broadcast along the spatial dimension and vice versa. F ′′ is the final
refined output. Figures 7 and 8 describe the computation process of each attention map. The details of each
attention module are described below.

Channel attention module. The structure of the channel attention module is shown in Fig. 7. The channel atten-
tion module aggregates spatial information of feature maps by using average pooling and max pooling opera-
tions, generating two different spatial context descriptors: Favg
c and F c , which denote average-pooled features
max
and max-pooled features respectively. The two descriptors are then forwarded to the shared network to generate
the channel attention map Mc ∈ RC×1×1. The shared network consists of a multilayer perceptron (MLP) with
one hidden layer. To reduce parameter overhead, the hidden activation size is set to RC/r×1×1, where r is the

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 8

Vol:.(1234567890)
www.nature.com/scientificreports/

Figure 8.  Diagram of spatial attention module.

reduction ratio. After applying the shared network to each descriptor, the channel attention module merges the
output feature vectors using element-wise summation. Briefly, the channel attention module is computed as:
Mc (F) = σ (MLP(AvgPool(F)) + MLP(MaxPool(F)))
  
c
   c  (3)
= σ W1 W0 Favg + W1 W0 Fmax

where σ represents the sigmoid function, W0 ∈ RC/r×C , and W1 ∈ RC×C/r . Note that both inputs share the MLP
weights W0 and W1 ,and the ReLU activation function is followed by W0.

Spatial attention module. The spatial attention module utilizes the spatial relationship of features to generate
a spatial attention map. Figure 8 depicts the computation progress of the spatial attention map. Unlike chan-
nel attention, spatial attention focuses on ‘where’ is an informative part, which is complementary to channel
attention. To compute the spatial attention, average pooling and max pooling operations are first applied along
the channel axis and then concatenated to generate an efficient feature descriptor. On the concatenated feature
descriptor, a convolution layer is applied to generate a spatial attention map Ms (F) ∈ RH×W , which encodes
locations to be emphasized or suppressed. The specific operation progress is as follows.
The channel information of the feature maps is aggregated using two pooling operations to generate 2D maps:
s ∈ R1×H×W and F s
Favg 1×H×W . Each represents the average-pooled and max-pooled features across the
max ∈ R
channel. They are then concatenated and convolved through standard convolutional layers to generate a 2D
spatial attention map. To sum up, the spatial attention is calculated as:
Ms (F) = σ f 7×7 ([AvgPool(F); MaxPool(F)])
 
   (4)
= σ f 7×7 F sang ; Fmax
s

where σ represents the sigmoid function and f 7×7 denotes a convolution operation with a filter size of 77.
The improved algorithm is based on SSD-ResNet-50 and adds CBAM modules after 6 feature maps of dif-
ferent sizes. By adjusting the size of the feature maps, each feature map can remain the same size as before the
output after the feature weighting of the CBAM module. The attention mechanism further enhances the semantic
information of high-level feature maps, reduces the target missed detection rate, and improves the robustness
of the algorithm.

Improvement of the loss function. The loss function of the original SSD network consists of a weighted
sum of the confidence loss and the position loss. The specific expression is calculated as:
1
(5)

L(x, c, l, g) = Lconf (x, c) + αLloc (x, l, g)
N
where Lconf and Lloc denote the confidence loss and localization loss, respectively. α represents the weight of
the localization loss, c and l are the category confidence and position offset information of the prediction box,
respectively. x is the matching result between the previous frame and different categories, if it matches, the result
is x = 1, otherwise, it is x = 0. Furthermore, g represents the offset between the ground-truth bounding box and
the prior bounding boxes, and N denotes the number of the prior bounding boxes.
The problem of sample imbalance can be solved by adjusting the positive and negative sample ratio param-
eters in the input network. To solve the problem of sample imbalance in the SSD algorithm, we use the Focal
Loss ­function22 to replace the confidence loss function in the original loss function. Its specific expression can
be summarized as:
γ
(6)
    
FL pt = −at 1 − pt log pt

where pt represents the classification probabilities of the different classes, and the value of γ is greater than zero,
which is used to adjust the rate at which the weights of the samples are easily divided. at is a decimal between 0
and 1, which is used as a weight to adjust the proportion of positive and negative samples. For simple samples,

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 9

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 9.  Curve of focal loss ­function22.

Figure 10.  Passive terahertz hidden object imaging. (a) The 0.2 THz passive terahertz imaging device. (b) the
metal pistol model and the mobile phone model. (c) the passive terahertz imaging results.

pt will be relatively large, and the weight will naturally be small. On the contrary, for difficult samples, pt will be
relatively small, and the weight will naturally be relatively large so that the network tends to use such difficult
samples to update parameters.
The Focal Loss is visualized for several values of γ ∈ [0, 5] in Fig. 9. The function curve is the loss function
curve in SSD when γ =0. It can be seen that even the loss function value of the easily separable samples is high,
resulting in a high proportion of the loss value of the easily separable samples in the algorithm. The weight of
the hard samples in the input samples increases with the increase of γ , indicating that Focal Loss achieves the
balance of the positive and negative samples, hard and easy samples through at and γ , respectively. In this way,
the samples participating in the training can be distributed more evenly, thereby further improving the reliability
of the detection algorithm. In this paper, we set the parameters at = 0.25 and γ = 2 during the training process,
because extensive experiments show that such parameters can achieve the best experimental results.

Experimental results and discussion

In this section, the image dataset is introduced first, and then the evaluation metrics are described. The results
of the ablation experiments demonstrate the effectiveness of the improved points proposed by this method.
Extensive comparison experiments between the improved SSD algorithm and other methods are conducted in
succession. In addition, we give an intuitive discussion about the visual detection effect of this algorithm and
depict the ROC curve.

Datasets. In the research field of terahertz concealed object detection, few passive image datasets are cur-
rently publicly available. The passive terahertz dataset we used in this paper was constructed by a team led by
researcher Chao Li from the Aerospace Information Research Institute, Chinese Academy of Sciences. The data-
set was collected by the 0.2 THz passive imaging device shown in Fig. 10a, a total of 896 images were collected.
The size of each image is 160392, and 0 2 objects to be detected (“pistol model” and “mobile phone model”) are
randomly hidden under clothes in different positions of the human torso. To ensure the authenticity and reli-
ability of the results, the imaging experiments were carried out in the form of real people carrying concealed
objects. The pistol model and mobile phone model carried by the volunteers are shown in Fig. 10b, and the
imaging results are shown in Fig. 10c. In the future, we plan to gradually increase the types of objects. In addi-
tion, 896 images are expanded to 1792 images by horizontal flipping through the data augmentation method to

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 10

Vol:.(1234567890)
www.nature.com/scientificreports/

Figure 11.  Definitions of true positive, false positive, false negative and true n
­ egative39. For example, “True
Positive” means that one actual positive instance was predicted to be positive by one method.

increase the diversity of samples. The following training experiments, evaluation experiments, and comparison
experiments are performed on this dataset.

Evaluation metrics. Since passive terahertz image object detection in the security field is still a relatively
“niche” field in computer vision, there is no standard evaluation protocol defined. We evaluate our proposed
improved SSD algorithm using mean precision (mAP), F1 score, ROC curve, AUC value, and FPS. The following
is a comprehensive overview of the indicators.
Object detection performance can be evaluated with Average Precision (AP) and mean Average Precision
(mAP) between ground truth and predicted bounding box (IOU). AP is defined as the area enclosed by the
precision and recall curves and the axes. The precision rate is the accuracy rate, which measures the classifica-
tion ability of the object detection algorithm to detect objects. The higher the accuracy rate, the stronger the
model’s ability to classify the object. The recall rate measures the detection ability of the model to the object. The
higher the recall rate, the stronger the model’s ability to distinguish the object. In the object detection algorithm,
the precision-recall curve (PR curve) of each category can be obtained. The area enclosed by the curve and the
abscissa axis is the average precision AP value. The mAP can be obtained by averaging the APs of all classes.
Under the same confidence, the larger the mAP of the model, the better the detection performance of the model.
The precision and recall rate are calculated as follows:
TP
Precision = (7)
TP + FP

TP
Recall = (8)
TP + FN
True positive (TP) means that the model predicts a positive sample and the actual prediction is correct;
false positive (FP) means that the model predicts a positive sample and detects it incorrectly, and false negative
(FN) means that the model predicts a negative sample and detects it incorrectly. These evaluation metrics are
illustrated in Fig. 11.
Additionally, the F1 score is a composite metric that combines precision and recall. The advantage of the
F1 score is that it provides a single measure of quality that is easier to understand. Its calculation formula is as
follows:
Precision × Recall
F1 Score = 2 × (9)
Precision + Recall
A receiver operating characteristic (ROC) ­curve40, commonly used to visualize the discriminant ability of
a binary classifier, is a plot of the true positive rate (TPR, sensitivity, recall) versus the false positive rate (FPR,
1-specificity) at various given cut points. A typical ROC curve is a plot of FPR on the x-axis versus TPR on the
y-axis. The area under the curve (AUC) maps the entire ROC curve into a single number that reflects the overall
performance of the classifier over all thresholds. Typically, the value of AUC is between 0.5 and 1. The larger
the value, the better the discriminative ability of the classifier. The formulas for calculating TPR and FPR are as
follows:
TP
TPR = (10)
TP + FN

FP
FPR = (11)
FP + TN
The detection speed is mainly used to evaluate the real-time performance of the algorithm. If the inference
speed cannot meet the real-time requirements, our model cannot be used in the real environment. FPS (frames
per second) is currently a common metric for evaluating model speed. In two-dimensional image detection,

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 11

Vol.:(0123456789)
 www.nature.com/scientificreports/

Loss function curve

25
total loss
reg loss
20 cls loss

15

Loss
10

0
0 1000 2000 3000 4000 5000 6000
Iteration

Figure 12.  Change of loss function.

Module mAP(%)
NET VGGNet-16 ResNet-50 Feature Fusion CBAM Focal Loss
√
SSD 95.04
√
Model-1 98.65
√ √
Model-2 98.77
√ √
Model-3 98.24
√ √
Model-4 95.37
√ √ √
Model-5 98.86
√ √ √ √
SSD(ours) 99.92

Table 2.  Ablation experiment results on the passive terahertz dataset.

it is defined as the number of pictures that can be processed per second. The larger the value, the shorter the
detection time and the faster the speed.

Experimental details. The experimental environment of this paper is: in terms of hardware, we adopt
the Inter(R)Core(TM)i5-10400F CPU processor and the graphics card of GeForce RTX 3090 Ti. In terms of
software, we choose the Ubuntu 20.04 operating system, Pytorch deep learning framework, and Python 3.6
programming language.
Batch_size is the number of images sent to training at one time. A larger batch_size can make model training
more stable, but it will increase the amount of computation. Considering the computing power of the graph-
ics card we have, we chose 16 as the batch_size. The learning rate is the speed at which the model is iteratively
trained, and setting the learning rate correctly can make the loss curve smoother. The initial learning rate was
set to 2e-4 and the learning rate decay weight was 0.001. The number of training iterations is 6000, the parameter
update method is the gradient descent method with momentum (momentum SGD), and the momentum fac-
tor is set to 0.9. The resulting model was saved after every 500 iterations in the training process. After training,
the object detection accuracy is used as the criterion for selecting the optimal model. The model is validated
using the test set, and the experimental results are analyzed. The change of the loss function during the training
process is shown in Figure 12. After 6000 iterations, the network converges. After 6000 iterations, the total loss
of the model can converge to 1.62.

Ablation study of the improved SSD algorithm. This section carries out the ablation study of the
improved SSD algorithm to prove its superiority over the baseline model. We analyze the impact of improving
feature extraction network, adding feature fusion network, adding attention mechanism, and improving loss
function on recognition accuracy. The results of the ablation experiments are shown in Table 2.
Model-1 replaces the backbone network from VGGNet-16 to ResNet-50, and the mAP is 3.61% higher than
that of SSD. Model-2 adds a feature fusion module on the basis of adding ResNet-50, and constructs a feature
with rich semantic information, indicating that mAP is 3.73% higher than SSD and 0.12% higher than Model-1.
Model-3 adds a space-channel attention mechanism CBAM based on adding ResNet-50, and the mAP is 3.20%
higher than that of SSD. Model-4 replaces the loss function with Focal Loss based on adding ResNet-50, and

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 12

Vol:.(1234567890)
www.nature.com/scientificreports/

Comparison of mAP values of different models

1
0.99
0.98
0.97
0.96
0.95
0.94
0.93

mAP
0.92
SSD(ours)
0.91
Model-1
0.9 Model-2
0.89 Model-3
0.88 Model-4
0.87 Model-5
SSD(300)
0.86
0.85
1000 2000 3000 4000 5000 6000
Iteration

Figure 13.  Comparision of mAP values of different models.

AP(%) mAP(%) F1 score Time(ms) FPS

Model Gun Phone
Faster R-CNN 98.30 97.20 97.75 0.95 9 4.5568
YOLO v5 99.24 97.48 98.36 0.96 42 12.2265
Retinanet 96.47 93.64 95.06 0.94 25 10.1952
SSD 96.10 93.98 95.04 0.93 23 11.2652
SSD(our) 99.94 99.90 99.92 0.98 17 8.2624

Table 3.  Comparison of accuracy with other advanced algorithms.

mAP is 0.33% higher than SSD. Model-5 replaces the feature extraction network with ResNet-50 on the basis
of the original SSD algorithm and adds a feature fusion module and a hybrid attention mechanism CBAM. The
results show that the mAP of Model-5 is 3.82% higher than that of SSD. Finally, the mAP of the fully improved
SSD algorithm can reach 99.92%. In summary, adding different modules significantly improves the performance
of the algorithm, and these gains add up to validate the effectiveness of our algorithm.
Furthermore, the comparison of mAP values obtained by different models in the first 6000 iterations is shown
in Fig. 13. It can be seen that our proposed improved model has the highest mAP value with a high mAP of
99.92% at the 6000th iteration.

Comparison with the state‑of‑the‑art detection methods. We evaluated the improved SSD algo-
rithm to other state-of-the-art detection approaches, including Faster RCNN, YOLO v5, RetinaNet, and the
baseline model SSD, to verify its comprehensive performance. Table 3 shows that the improved SSD model has
a detection accuracy of 99.92%. It has the best accuracy, even outperforming the existing YOLO v5 detection
model. Although the improved SSD model is slower than the fastest YOLO v5, it offers a 1.56% accuracy gain. At
the same time, the improved model has a speed of 17 frames per second, which is sufficient for real-time detec-
tion. The aforementioned findings suggest that our proposed enhanced algorithm outperforms other techniques.
As shown in Table 3, the detection accuracy of the improved algorithm is improved compared with the
mainstream algorithms. Meanwhile, due to the increased complexity of the network, the algorithm in this paper
sacrifices the speed, but it can still meet the needs of real-time detection. We believe there are two primary
reasons for such superiority. (1) The baseline model of the improved SSD algorithm is suitable for this task. Its
detection accuracy is only 2.71% lower than Faster RCNN, but the speed is 2.55 times faster. (2) The improved
SSD algorithm can robustly improve the accuracy of the baseline model. The specific reasons can be seen in
“Ablation study of the improved SSD algorithm”.
We employed the passive terahertz human security image dataset given in the “Datasets” section above for
detection experiments, and the visual detection results are shown in Fig. 14. As demonstrated in the red boxes in
Fig. 14a, the original SSD algorithm has missed detection, and the detection confidence also has to be enhanced.
The improved SSD algorithm effectively addresses the issue of missed detection while also enhancing detection
confidence. The improved algorithm’s detection results are displayed in Fig. 14b. To summarize, the improved SSD
algorithm outperforms the original algorithm in terms of robustness and accuracy, as well as overall performance.

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 13

Vol.:(0123456789)
 www.nature.com/scientificreports/

(a)

(b)

Figure 14.  Visualized results of the passive terahertz dataset. (a) Detection results of the original SSD
algorithm. (b) Detection results of the improved SSD algorithm.

Figure 15.  ROC curve.

ROC curve. The ROC curve of the proposed model is shown in Fig. 15, demonstrating the excellent recogni-
tion ability of our model for concealed objects in passive terahertz images. AUC provides summary measures
for the performance of a system. It provides meaningful performance analysis. The AUC value of our proposed
model is 0.87, indicating that the model has a good discriminative ability.

Conclusion
To further improve the detection accuracy and speed of concealed objects in terahertz human security images, we
proposed an improved SSD algorithm to promote its algorithm performance in this paper. First, the ResNet-50
network was used to replace the original VGGNet-16 network to overcome the VGG network degradation
problem. Afterward, we designed a feature fusion module to fuse deep features and shallow features to construct
features rich in semantic information, which is beneficial to the detection of small objects. In the next parts, the
hybrid attention mechanism was introduced in the SSD network to improve the network’s attention to concealed
objects. Finally, we introduced the Focal Loss function in the loss function to improve the robustness of the
algorithm. The results of ablation experiments illustrate the efficacy of the changes described in this paper. At

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 14

Vol:.(1234567890)
www.nature.com/scientificreports/

the same time, we also compare the proposed model with other mainstream detection methods on the terahertz
human security image dataset. The results showed that our method achieves significantly improved detection
accuracy in comparison with the original algorithm when the speed is only slightly reduced. The detection
accuracy of the proposed method is as high as 99.9%, and the detection speed is about 17 FPS. Therefore, it can
meet the real-time detection needs of security inspection scenarios.
Our approach can be improved in the future as follows: (1) The training dataset can be augmented to improve
the generalization ability of the model. (2) The model can be further lightweight to reduce the amount of cal-
culation and improve the detection speed. (3) More emphasis should be placed on the model’s interpretability
in order to make the model easier to interpret and understand. It also makes the black box of the CNN-based
model more transparent and reasonable. We believe that the application of Explainable AI will help to reduce
the uncertainty of deep learning models. The insights gained from this research facilitate the application of deep
learning techniques in smart security screening scenarios. It is believed that the method proposed in this paper
will have important application value in the terahertz intelligent security systems.

Data availability
The datasets generated during the current study are not publicly available to non-cooperative units in the short
term due to the dataset has been collected for a long time and the data involves personal privacy. However, the
dataset is available from the corresponding author upon reasonable request.

Received: 16 April 2022; Accepted: 6 July 2022

References
1. Luo, J. J., Hou, S. X., Jun, X. U. & Juan-Juan, L. I. Study on applications of terahertz technology. Laser Ultrav. Terahertz Technol.
6622, 29–35 (2008).
2. Appleby, R. & Wallace, H. B. Standoff detection of weapons and contraband in the 100 ghz to 1 thz region. IEEE Trans. Antennas
Propag. 55, 2944–2956 (2007).
3. Erich, et al. Passive terahertz camera for standoff security screening. Appl. Opt. 49, E106-20 (2010).
4. Jiang, Y., Ge, H., Lian, F., Zhang, Y. & Xia, S. Early detection of germinated wheat grains using terahertz image and chemometrics.
Sci. Rep. 6, 21299 (2016).
5. Karaliūnas, M. et al. Non-destructive inspection of food and technical oils by terahertz spectroscopy. Sci. Rep. 8, 1–11 (2018).
6. Khaliduzzaman, A., Konagaya, K., Suzuki, T., Kashimori, A. & Ogawa, Y. A nondestructive eggshell thickness measurement
technique using terahertz waves. Sci. Rep. 10, 1052 (2020).
7. Guo, L. T., Deng, C., Zhao, Y. M. & Zhang, C. L. Passive terahertz imaging for security application. In ISPDI 2013 - Fifth Interna-
tional Symposium on Photoelectronic Detection and Imaging (2013).
8. Ran, Z., Zhao, Y. M., Chao, D., Zhang, C. L. & Li, Y. Target recognition in passive terahertz image of human body. Proceedings of
Spie the International Society for Optical Engineering9275 (2014).
9. Pun, T. A new method for gray-level picture thresholding using the entropy of the histogram. Signal Process. 2, 223–237 (1985).
10. Ostu, N. A threshold selection method from gray-histogram. IEEE Trans. Syst. Man Cybern. 9, 62–66 (1979).
11. Bengio, et al. Representation learning: A review and new perspectives. IEEE Trans. Pattern Anal. Mach. Intell. 35, 1798–1828
(2013).
12. Dalal, N. & Triggs, B. Histograms of oriented gradients for human detection. In IEEE Computer Society Conference on Computer
Vision & Pattern Recognition (2005).
13. Lowe, D. G. Object recognition from local scale-invariant features. In Proc of IEEE International Conference on Computer Vision
(1999).
14. Zhang, S., Wen, L., Bian, X., Lei, Z. & Li, S. Z. Single-shot refinement neural network for object detection. In 2018 IEEE/CVF
Conference on Computer Vision and Pattern Recognition (2018).
15. Tian, Z., Shen, C., Chen, H. & He, T. Fcos: Fully convolutional one-stage object detection. In 2019 IEEE/CVF International Confer-
ence on Computer Vision (ICCV) (2020).
16. Liu, C., Yang, M. H. & Sun, X. W. Towards robust human millimeter wave imaging inspection system in real time with deep learn-
ing. Prog. Electromagn. Res. 161, 87–100 (2018).
17. Xiao, H., Zhang, R., Wang, H., Zhu, F. & Zhou, Y. B. R-pcnn method to rapidly detect objects on thz images in human body security
checks. In 2018 IEEE SmartWorld, Ubiquitous Intelligence and Computing, Advanced and Trusted Computing, Scalable Computing
and Communications, Cloud and Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/
ATC/CBDCom/IOP/SCI) (2018).
18. Yang, X. et al. Cnn with spatio-temporal information for fast suspicious object detection and recognition in thz security images.
Signal Process. 160, 202–214 (2019).
19. Simonyan, K. & Zisserman, A. Very deep convolutional networks for large-scale image recognition. Computer Science (2014).
20. He, K., Zhang, X., Ren, S. & Sun, J. Deep residual learning for image recognition. CoRRabs/1512.03385 (2015). ArXiv:1​ 512.0​ 3385.
21. Lin, T. Y. et al. Feature pyramid networks for object detection. IEEE Computer Society (2017).
22. Lin, T.-Y., Goyal, P., Girshick, R., He, K. & Dollár, P. Focal loss for dense object detection. IEEE Trans. Pattern Anal. Mach. Intell.
42, 318–327. https://​doi.​org/​10.​1109/​TPAMI.​2018.​28588​26 (2020).
23. Girshick, R., Donahue, J., Darrell, T. & Malik, J. Rich feature hierarchies for accurate object detection and semantic segmentation.
IEEE Computer Society (2013).
24. Girshick, R. Fast r-cnn. In Proceedings of the IEEE international conference on computer vision, 1440–1448 (2015).
25. Ren, S., He, K., Girshick, R. B. & Sun, J. Faster R-CNN: towards real-time object detection with region proposal networks. CoR-
Rabs/1506.01497 (2015). ArXiv:​1506.​01497.
26. He, K., Gkioxari, G., Dollár, P. & Girshick, R. B. Mask R-CNN. CoRRabs/1703.06870 (2017). ArXiv:​1703.​06870.
27. Dai, J., Li, Y., He, K. & Sun, J. R-fcn: Object Detection via Region-based Fully Convolutional Networks (Curran Associates Inc., New
York, 2016).
28. Redmon, J., Divvala, S. K., Girshick, R. B. & Farhadi, A. You only look once: Unified, real-time object detection. CoR-
Rabs/1506.02640 (2015). ArXiv:​1506.​02640.
29. Liu, W. et al. SSD: single shot multibox detector. In Leibe, B., Matas, J., Sebe, N. & Welling, M. (eds.) Computer Vision - ECCV
2016 - 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part I, vol. 9905 of Lecture Notes
in Computer Science, 21–37, https://​doi.​org/​10.​1007/​978-3-​319-​46448-0_2 (Springer, 2016).

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 15

Vol.:(0123456789)
 www.nature.com/scientificreports/

30. Morales, P., PeRez-Suay, A., Molina, R., Camps-Valls, G. & Katsaggelos, A. K. Passive millimeter wave image classification with
large scale gaussian processes. In 2017 IEEE International Conference on Image Processing (ICIP) (2018).
31. Fang et al. Study of automatic detection of concealed targets in passive terahertz images for intelligent security screening. IEEE
Transactions on Terahertz Science and Technology (2019).
32. Zhang, X., Zhao, Y. & Zhang, C. Passive terahertz image segmentation algorithm. High power Laser and Particle Beam25 (2013).
33. Jiang, T., Shen, H., Yang, D., Liu, J. & Zou, Z. Target detection of thz images based on c-means of fuzzy local information. Laser
Technol. (2015).
34. Tapia, S. L., Molina, R. & de la Blanca, N. P. Detection and localization of objects in passive millimeter wave images. In 24th Euro-
pean Signal Processing Conference, EUSIPCO 2016, Budapest, Hungary, August 29 - September 2, 2016, 2101–2105, https://​doi.​org/​
10.​1109/​EUSIP​CO.​2016.​77606​19 (IEEE, 2016).
35. Niu, Y., Wang, Z., Qiao, L. & Zhao, Z. Automatic target recognition method for millimeter-wave body scanner. In 2016 IEEE
International Conference on Microwave and Millimeter Wave Technology (ICMMT) (2016).
36. Hinton, G. E. & Salakhutdinov, R. R. Reducing the dimensionality of data with neural networks. Science313.
37. Ma, Y., Zhang, P. & Tang, Y. Research on fish image classification based on transfer learning and convolutional neural network
model. In 2018 14th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD)
(2018).
38. Woo, S., Park, J., Lee, J. & Kweon, I. S. CBAM: convolutional block attention module. In Ferrari, V., Hebert, M., Sminchisescu, C. &
Weiss, Y. (eds.) Computer Vision - ECCV 2018 - 15th European Conference, Munich, Germany, September 8-14, 2018, Proceedings,
Part VII, vol. 11211 of Lecture Notes in Computer Science, 3–19, https://​doi.​org/​10.​1007/​978-3-​030-​01234-2_1 (Springer, 2018).
39. Schiele, et al. Pedestrian detection: An evaluation of the state of the art. IEEE Trans. Pattern Anal. Mach. Intell. 34, 743–761 (2012).
40. Fawcett, T. An introduction to roc analysis. Pattern Recogn. Lett. 27, 861–874 (2005).

Acknowledgements
This work was supported by the Key-Area Research and Development Program of Guangdong Province
(2019B010157001, 2020B0101110001); the National Natural Science Foundation of China (61988102,
61731020, 61671432); the National Key Research and Development Program of China (2017YFA0701004,201
8YFF01013004); the Project of Equipment Pre-Research (WJ2019G00019); and the Key Program of Scientific
and Technological Innovation from Chinese Academy of Sciences (KGFZD-135-18-029).

Author contributions
LC performed the theoretical study, conducted the experiments, processed the data, and wrote the manuscript.
YJ analyzed the results and revised the manuscript. CL designed the imaging system and revised the manuscript.
XL and GF provided the experiment equipment and funds for the research. All authors have read and agreed to
the published version of the manuscript.

Competing interests
The authors declare no competing interests.

Additional information
Correspondence and requests for materials should be addressed to Y.J.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.

© The Author(s) 2022

Scientific Reports | (2022) 12:12082 | https://doi.org/10.1038/s41598-022-16208-0 16

Vol:.(1234567890)