A Review of Different Segmentation Approach on Non Uniform Images

The segmentation approach plays an important role in image processing, especially for detection and identification. However, a poor image quality causes a shadow, artifacts, and non-uniform background will reduce the segmentation effectiveness. This article provides a comprehensive study of a few segmentation techniques such as Otsu Method, Double Mean Value (DMV) method, Gradient Based Thresholding, Yanowitz and Bruckstein's (YB) method, Chen's method, Blayvas's method, Chan's method and Niblack's method. The objective of this study is to explore the mathematical algorithm and performing of each segmentation methods. In order to evaluate the performance, the Misclassification Error (ME) was obtained. The overall results of the numerical simulation indicate that the Gradient Based method achieved 0.0199 and followed by Chen method 0.0226.