Digital Image Segmentation

Abstract:

Image segmentation plays an important role for many image video and computer vision applications. It is still a relevant research area due to its wide usage in the field of medical, remote sensing and image retrieval. Image segmentation is typically used to locate objects and boundaries in images. It is used to classify or cluster an image into several parts (regions) according to the feature of image, for example, the pixel value or the frequency response.
Several image segmentation algorithms were proposed to segment an image before recognition or compression. This paper focuses towards highlighting the strength and limitations of earlier proposed classification techniques discussed in the contemporary literature. This paper

Image segmentation is very important and challenging process of image processing. It is an essential step in image analysis, object representation, visualization, and many other image processing tasks. It is the technique that divides the image into different segments. The different segmentation Techniques are

Threshold based segmentation:
In this method, image pixels are divided with the help of intensity level of an image. This method is mainly used to distinguish the foreground objects from background images. Histogram thresholding and slicing techniques are used to segment the image.
If more than two segments are required, the method described above can be extended to use multiple thresholds. If the number of segments is large, a more practical algorithm that minimizes the variances within segments is often used, an iterative algorithm known as K-means clustering.
Edge based segmentation:
In this technique, detected edges in an image are assumed to represent object boundaries, and they are used to identify these objects.
A typical approach to segmentation using edges is (1) compute an edge image, containing all edges of an original image, (2) process the edge image so that only closed object boundaries remain, and (3) transform the result to an ordinary segmented image by filling in the object

Region growing needs a set of starting pixels called seeds. The region growing process consists of picking a seed from the set, investigating all 4-connected neighbors of this seed, and merging suitable neighbors to the seed. The seed is then removed from the seed set, and all merged neighbors are added to the seed set. The region growing process continues until the seed set is empty.
Texture measures:
Clustering techniques: clustering methods attempt to group together patterns that are similar in some sense. These techniques can be divided into three categories: processing of the original image prior to segmentation, processing of the segmented result, or adaptation of the segmentation process.
Agglomerative clustering. This approach starts out by calling each data point a separate cluster, and then proceeds to merge appropriate clusters into single clusters. The dissimilarity matrix is used to decide which clusters are to be merged; the smallest entry gives the data points that are least dissimilar and hence are the most likely candidates to be merged. After merging, the next smallest entry is selected, supplying two new merging candidates,