IRJET- A Survey on Object Detection using Deep Learning Techniques
0 likes55 views
1) The document discusses object detection and tracking using deep learning techniques such as convolutional neural networks, YOLO, and Single Shot Detector. It reviews literature on existing approaches and proposes a system for multi-purpose security applications using mobile camera detection. 2) Common object detection methods discussed are Faster R-CNN, YOLO, and Single Shot Detectors (SSDs). The proposed system would use these deep learning techniques for automated detection and tracking from mobile cameras. 3) Applications mentioned include surveillance, people counting, drowsiness detection, and more. The growth of mobile phones makes them suitable for real-time detection and tracking technologies.
IRJET- A Survey on Object Detection using Deep Learning Techniques
- 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 06 Issue: 02 | Feb 2019 www.irjet.net p-ISSN: 2395-0072 © 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 2140 A Survey on Object Detection using Deep Learning Techniques G. Kishore1, G. Gnanasundar2, Sri. Harikrishnan3 B.E. Students, Department of Computer Science and Engineering, SRM Valliammai Engineering College. -------------------------------------------------------------------------***------------------------------------------------------------------------ Abstract - Object detection and tracking will be done usually by surveillance camera or any kind of sensors. We are going to do detection and tracking using surveillance camera and also mobile camera. Mobile phones have become handy and improved in many aspects, so that detection and tracking in mobile camera will be useful. This project will mostly depend on deep learning algorithm. Algorithm that is specifically used in this project will be convolution neural network. We willdevelopauserinterfaceto access the detection of objects as well as tracking. Key Words: Object detection, Object tracking, Deep learning, Surveillance camera, Mobile camera 1. INTRODUCTION Object detection and tracking has become a major part in today’s technology. Object detection is the process of finding the instances of real-world objects such as faces, bicycles, buildings and many real time objects. Object detection is the task of detecting the object and drawing a bounding box around them, i.e. localizing them. Object tracking is used to track the particular object or many objects at same instance in the given camera range. Real time object detection and tracking has played an important role in major researches because of its increased commercial applications such as surveillance system, Mobile robots, Border security, People counting, Wild animals monitoring and many more. 1.1Object Tracking: Object tracking is the process of locating a moving object over time using a camera. Object tracking can be a time consuming process due to the amount of data that is contained in video. Object tracking can be done by splitting a video source into multiple sub frames. 1.2Object Detection: Object detection is the process of finding the instances of real-world objects such as faces, bicycles, buildings and many real time objects. Object detection is the task of detecting the object and drawing a bounding box around them, i.e. localizing them. 1.3Face Recognition: Face recognition is a widely used technology in various security applications. Face recognition is the basic technology used in emotion recognition and also used to verify a person’s authenticity. 2. LITERATURE SURVEY S.NO TITLE AUTHOR CONCEPT YEAR 1 Application of deep learning in object detection Xinyi Zhou, Wei Gong, WenLong Fu, Fengtong Du DATASETS: ImageNet PASCALVOC COCO METHODOLOGY: R-CNN SPP-net Fast R-CNN Faster R-CNN 978-1-5090-5507- 4/17/$31.00 ©2017 IEEE ICIS 2017, May 24-26, 2017, Wuhan, China 2017 2 An Efficient approach for object detection and object tracking B. Maga MODULE: Kernel Method and Training Feature based tracking method 2017 Third International Conference on Science Technology Engineering and Management
- 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 06 Issue: 02 | Feb 2019 www.irjet.net p-ISSN: 2395-0072 © 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 2141 Template Generation TECHNIQUE: Template Matching Later Based Tracking 3 Object detection basedon deep learning of small samples Ce Li, Yachao Zhang KEY STEPS: Foreground objects extraction Background selectionand fusion processing Object Semantic-relevant Detection 2018 Tenth International Conference on Advanced Computational Intelligence (ICACI) March 2018, Xiamen, China 4 A Learning algorithm for model-based object detection Chen Guodong, Zeyang Xia, Rongchuan Sun, Zhenhua Wang, Zhiwu Ren and Lining Sun KEY STEPS: Object detection Shape Matching Image Segmentation Shape Fragment PROPERTIES: Rotation invariance Scale invariance Noise robustness 2011,8th International Conference on Ubiquitous Robots and Ambient Intelligence 5 Object detection and tracking K. Rasool Reddy, K. Hari Priya, N. Neelima KEY STEPS: Incremental Multiple principle component analysis Frag Track HOG - LBP Detector Generative and Discriminative Trackers Semi Supervised Support Vector Machines 2015 INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCDE AND COMMUNICATION NETWORK 6 Modelling from an object and multi-object tracking system Afef SALHI, Yacine MORESLY,Fahmi GHOZZI, Ameni YENGUI, and Ahmed FAKHFAKH KEY STEPS: Block-matching KLT algorithm (Kanade Lucas Tomasi) Meanshift algorithm(MA) Camshift Algorithm (CA) 2016, Global Summit on Computers and Information Technology 7 Object detectioninsports video M. Buric, M. Pobar, M.Ivasic- Kos METHODOLOGY : Mask R-CNN YOLO object detector Mixture of Gaussians method MIPRO 2018, May 21-25, 2018, Opatija Croatia 8 Object Tracking Camera Priyanka Pacharne, Sanket Kotkar, Neha Darekar KEY STEPS : Colour Model Object Tracking Image Acquisition Background Subtraction IJSRD - International Journal for Scientific Research & Development| Vol. 3, Issue 03, 2015 | ISSN (online): 2321- 0613
- 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 06 Issue: 02 | Feb 2019 www.irjet.net p-ISSN: 2395-0072 © 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 2142 9 A Survey on Object Tracking in Video Snehlata Raisagar, Ashish Tiwari KEY STEPS : Video Sequence Object Detection Object Recognition Tracking 2017, IJSRD - International Journal for Scientific Research & Development 10 DetectionandTrackingof Moving Object in Video - A Survey Dhaval Deshpande,Nikhil Aatkare, Prof.Reena Somani STATISTICAL METHODS : Background Subtraction Temporal Differencing Correspondence Based Matching Algorithm Kernel Tracking 2016, || National Conference on Technological Advancement and Automatization in Engineering 3. PROPOSED SYSTEM: In our proposed project, we are going to provide multi-purpose security system that can be used in various applications like surveillance security in military, animal monitoring,peoplecountingincrowdedplaces,drowsinessdetection etc., Thisprocess can be done using deep learning techniques like convolution neural network, YOLO framework and Single Shot Detector . The usage of technique is based on the needed application. Entire project has been developed in OpenCV platform and python programming language. 3.1Object Detection Methods: • Faster R-CNNs Faster R-CNNs is the method for object detection using deep learning, even with the faster implementation R-CNNs (where the “R” stands for “Region Proposal”) the algorithm can be quite slow, on the order of 7 FPS. • You Only Look Once (YOLO) YOLO has the faster implementation capable of processing 40-90 FPS on a Titan X GPU. The fast variant of YOLO can even get up to 155 FPS. • Single Shot Detectors (SSDs) SSDs, originally developed by Google, are a balance between the two. The algorithm is more straightforward than Faster R-CNNs. SSDs also tend to be more accurate than YOLO. 4. CONCLUSION In our project we have created a simple way of using the object detection and tracking. This detectionandtracking will befully automated. Purpose of doing the detection and tracking using mobile camera is because of growth of the mobile phone technology. Statistics show that the number of mobile phone users in India has risen from 524.9 million to 813.2 million from the year of 2013 to 2019. So, providing the detection and tracking technology using mobile phone will have a great part in developing technology. REFERENCES 1. Xinyi Zhou, Wei Gong, WenLong Fu, Fengtong Du. Application of deep learning in object detection. 978-1-5090-5507- 4/17/$31.00 ©2017 IEEE ICIS 2017, May 24-26, 2017, Wuhan, China 2017. 2. B. Maga. An Efficient approach for object detection and object tracking. Third International Conference on Science Technology Engineering and Management 3. Ce Li, Yachao Zhang. Object detection based on deep learning of small samples. 2018 Tenth International Conference on Advanced Computational Intelligence (ICACI)March 2018, Xiamen, China 4. Chen Guodong, Zeyang Xia, RongchuanSun,Zhenhua Wang,ZhiwuRenandLining Sun. ALearningalgorithmformodel based object detection. 2011, 8th International Conference on Ubiquitous Robots and Ambient Intelligence.
- 4. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 06 Issue: 02 | Feb 2019 www.irjet.net p-ISSN: 2395-0072 © 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 2143 5. K. Rasool Reddy, K. Hari Priya, N. Neelima. Object detection and tracking. 2015 INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCDE AND COMMUNICATION NETWORK 6. Afef SALHI, Yacine MORESLY, Fahmi GHOZZI,Ameni YENGUI, and Ahmed FAKHFAKH. Modelling from an object and multi-object tracking system. 2016, Global Summit on Computers and Information Technology. 7. M. Buric, M. Pobar, M. Ivasic-Kos. Object detection in sports video. MIPRO 2018, May 21-25, 2018, Opatija Croatia. 8. Priyanka Pacharne, Sanket Kotkar, Neha Darekar. Object Tracking Camera. IJSRD - International Journal forScientific Research & Development| Vol. 3, Issue 03, 2015 | ISSN (online): 2321-0613. 9. Snehlata Raisagar, Ashish Tiwari. A Survey on Object Tracking in Video. 2017, IJSRD - International Journal for Scientific Research & Development. 10. Dhaval Deshpande, Nikhil Aatkare, Prof.Reena Somani. Detection and Tracking of Moving Object in Video - A Survey. 2016, || National Conference on Technological Advancement and Automatization in Engineering.