Scribd
Upload
28 views7 pages

Image Processing

The document discusses image segmentation techniques for partitioning digital images into segments. It describes the goals of image segmentation as detection of discontinuities, boundary detection, and region-based segmentation. It also explains different types of segmentation including contextual segmentation methods like region growing and split-merge segmentation, as well as non-contextual methods like thresholding.

Uploaded by

harsh dev
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
28 views7 pages

Image Processing

The document discusses image segmentation techniques for partitioning digital images into segments. It describes the goals of image segmentation as detection of discontinuities, boundary detection, and region-based segmentation. It also explains different types of segmentation including contextual segmentation methods like region growing and split-merge segmentation, as well as non-contextual methods like thresholding.

Uploaded by

harsh dev
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

Madhav Institute Of Technology And Science

Image Processing
Proficiency Presentation
Image Segmentation

Submitted by: Submitted to:


Harshdev Tripathi Dr. Vibha Tiwari
(0901AI211033)
Introduction to Image
Segmentation
Image segmentation is the process of partitioning a digital image into multiple
segments, or sets of pixels. This technique is crucial for object detection and boundary
identification, enabling advanced computer vision applications.

The primary goals of image segmentation include:

1.Detection of Discontinuities: Identifying significant changes like edges, lines, or points.


2.Boundary Detection: Outlining object boundaries for precise delineation.
3.Region-based Segmentation: Grouping pixels or regions with similar properties to
form coherent segments.
Discontinuities detection
Imagine an image as a landscape of brightness and color. Edge detection is like finding the cliffs
and ridges in this landscape. It helps us identify the places where things sharply change,
revealing the shapes and boundaries of objects.

There are different ways to find these changes:


•Spotting peaks: Like finding the lone, tall mountain – pinpointing single points with dramatic
shifts in brightness.
•Following lines: Imagine tracing rivers or canyons – identifying straight or curved lines where
intensity changes flow.
•Highlighting borders: This is like drawing lines around countries on a map – detecting edges
where regions with different brightness levels meet abruptly.
Types of Segmentation
1.Contextual Segmentation

Pixel Connectivity:
•Pixel connectivity refers to how pixels are considered connected or adjacent to each other in an image. It
determines the neighborhood relationships used in various image processing and segmentation algorithms.

Region Growing:
•Region growing is a segmentation technique where pixels or regions with similar properties are grouped
together to form larger segments.

Split-and-Merge Segmentation:
•Split-and-merge segmentation is a region-based segmentation method that involves dividing regions into
smaller segments (splitting) based on certain criteria, such as uniformity or homogeneity, and then merging
similar neighboring segments (merging) to create larger and more coherent regions.

Region Similarity:
•Region similarity refers to the measure of similarity or homogeneity between pixels or regions in an image.
•Similarity criteria can include intensity, color, texture, spatial proximity, or any combination of these
factors.
Types of Segmentation
2.Non-Contextual Segmentation

Simple Thresholding:
•Simple thresholding is a basic technique used to separate objects or regions
of interest from the background in an image based on pixel intensity.
•You choose a threshold value, and then each pixel in the image is classified
as either belonging to the foreground (object of interest) or the background
based on whether its intensity value is above or below the threshold.

Adaptive Thresholding:
•What is it? Adaptive thresholding is a more advanced technique that adjusts
the threshold value for different regions of an image based on local
characteristics.
•How does it work? Instead of using a single global threshold for the entire
image, adaptive thresholding calculates thresholds locally for small regions.
This is useful for images with varying lighting conditions or when objects
have different contrast levels.
Applications of Image
Segmentation in Object Detection
• Medical Imaging Segmenting MRI, CT, and ultrasound images to identify organs,
tumors, and other anatomical structures.

• Autonomous Vehicles Detecting and distinguishing pedestrians, other vehicles, traffic signs
and obstacles for safe navigation.
.
• Video Analysis and Surveillance: Moving Object Detection: Detecting and tracking moving
objects in video surveillance for security and traffic monitoring.

• Biometrics and Security: Face Recognition: Identifying and verifying individuals in facial
biometrics for security systems and access control.

• Document Processing: Text Extraction: Segmenting text regions in scanned documents for optical
character recognition (OCR) and document digitization.
Thank You

You might also like