Chapter Preview
TopIntroduction
Image (and video) segmentation is an important image technique, and is often described as the process that subdivides an image (or a clip of frames) into its constituent parts and extracts those parts of interest (objects). It is well known by its utility, since for extracting the useful information from images or a group/sequence of images, to separate the objects from background is an inevitable step/task. It is also well known by its complexity, as there is no general theory for image segmentation, yet. Therefore, the development of image segmentation techniques is still an ad hoc process.
Image segmentation is one of the most critical tasks in automatic image analysis, which is at the middle layer of image engineering (IE). Image Engineering (which is composed of three layers from bottom to top: image processing, image analysis and image understanding) is a new discipline and a general framework for all image techniques (Zhang, 2008d).
According to the statistics gathered from a yearly bibliography survey on image engineering (Zhang 2013), the journal publication on image segmentation is ranked the first among the current 16 groups/branches of research techniques of image engineering. The comprehensive survey has been made consecutively for 18 years, and the totally involved papers are more than 40000, in which 8243 are related to the different technique groups of image engineering. The statistics for the distribution of these papers in each group are listed in Table 1. It is seen that the group of image segmentation is the one that attracts the most attentions and achieve the most results among a complete list of technique groups.
Table 1. Journal papers in different technique groups
| No | Technique Group | # of Papers | No | Technique Group | # of Papers |
| 1 | Segmentation and edge detection | 1238 | 9 | Content-based image retrieval | 347 |
| 2 | Enhancement and filtering | 974 | 10 | Reconstruction from projections | 303 |
| 3 | Coding / decoding | 896 | 11 | Analysis and feature measurement | 287 |
| 4 | Object extraction and recognition | 832 | 12 | Object representation and description | 233 |
| 5 | Registration, matching and fusion | 810 | 13 | 3-D modeling and scene recovery | 231 |
| 6 | Biometrics | 643 | 14 | Multiple-resolution processing | 158 |
| 7 | Watermarking, and information hiding | 599 | 15 | Spatial-temporal technology | 90 |
| 8 | Capturing and storage | 523 | 16 | Image perception and interpretation | 79 |
Key Terms in this Chapter
Region Growing: Region growing is a region-based sequential technique for image segmentation by assembling pixels into larger regions based on predefined seed pixels, growing criteria and stop conditions.
Image Segmentation: A process consists of subdividing an image into its constituent parts and extracting these parts of interest (objects) from the image.
Graph Search: Graph search is a particular type of segmentation techniques which combing edge detection and linking together. It represents edge segments in the form of a graph and searching the graph for low-cost paths that correspond to significant edges or boundaries of objects.
Watersheds: Watershed technique is inspired from the topographic interpretation of images Segmentation by watersheds embodies many concepts of edge detection, thresholding and region processing techniques, and often produces stable and continuous results.
Clustering: Clustering is also called unsupervised learning and is a powerful technique for pattern classification. It is a process to group, based on some defined criteria, two or more terms together to form a large collection. In the context of image segmentation, it is often considered as the multi-dimensional extension of the thresholding technique.
Edge Detection: Edge detection is the most common approach for detecting discontinuities in images, and is the fundamental step in edge-based parallel process for segmentation. An edge is a local concept. To form a complete boundary of an object, edge detection should be followed by edge linking or connection.
Gradient Operator: Gradient operator is the first type of operators used for edge detection. The gradient of an image is a vector consisting of the first-order derivatives (including the magnitude and direction) of an image.
Thresholding: Thresholding techniques are the most popularly used segmentation techniques. A set of suitable thresholds need to be first determined, and then the image can be segmented by comparing the pixel properties with these thresholds.
Active Contour Model: Active contour model is a sequential technique for image segmentation. Given an approximation of the boundary of an object in an image, an active contour model can be used to find the “actual” boundary by deforming the initial boundary to lock onto features of interest within in this image.
Image Engineering: An integrated discipline/subject comprising the study of all the different branches of image and video techniques. It mainly consists of three levels: Image Processing, Image analysis, Image understanding.