Techniques for Evaluation of Image Segmentation Performance

Image segmentation is the entry to image analysis, so the result of image analysis depends greatly on type of segmentation technique used. Several techniques are used for segmentation of an image. For the judgement of quality of a segmented image several methods have been evolved. These methods evaluate the performance of the segmented image and provide a way for the selection of an optimal technique for image segmentation. In this paper the performance evaluation techniques for different segmentation methods have been discussed. These techniques provide a way for the empirical and theoretical evaluation of the available segmentation methods. Within these evaluation methods further sub-classes have been used according to the different image parameters, available segmentation evaluation algorithms and intuitive qualities of human nature.