Implementation and Comparison of different segmentation algorithms used for underwater images based on nonlinear objective assessments

The quality of underwater images is directly affected by water medium, atmosphere medium, pressure and temperature. This emphasizes the necessity of image segmentation, which divides an image into parts that have strong correlations with objects to reflect the actual information collected from the real world. Image segmentation is the most practical approach among virtually all automated image recognition systems. Feature extraction and recognition have numerous applications on telecommunication, weather forecasting, environment exploration and medical diagnosis. Different segmentation techniques are available in the literature for segmenting or simplifying the underwater images. The performance of an image segmentation algorithm depends on its simplification of image. In this paper, different segmentation algorithms namely, edge based image segmentation, adaptive image thresolding, watershed segmentation, Region growing segmentation, Quadtree segmentation and fuzzy c-means segmentation are implemented for bench mark images and they are compared using nonlinear assessment or the quantitative measures like gray level energy, discrete entropy, relative entropy, mutual information, normalized mutual information and redundancy. Out of the above methods the experimental results show that fuzzy c means clustering algorithm performs better than other methods in processing underwater images.