Teaching Tactics for Color Image Segmentation Using Otsu and Minimum Cross Entropy

The most fundamental step in image processing is image segmentation and it results in revealing enormous information embedded in most widely used RGB color space image. Excellent result is obtained for bi-level thresholding and the exhaustive search for optimal threshold values, to analyze complex images in multilevel thresholding (MLT), is reduced by promising objective functions such as Otsu and minimum cross entropy MCE aided with teaching-learning based optimization metaheuristic algorithm (TLBO). In TLBO, a teacher shares cognizance to a student. The use of only common control parameters and less-specific control parameters in TLBO achieves exploration and exploitation. The efficiency of TLBO is compared with cuckoo search algorithm (CS) at 4,5,6 and 7 threshold levels. Experimental results reveal that optimal output of TLBO is more successful in precise image segmentation and aids in various real time applications.