Specular Reflection Detection and Substitution: A Key for Accurate Medical Image Analysis

The quality of any image depends on the specifications of capturing devices. However, external factors also affect the appearance of an image. The disturbance created in image due to reflections from the surface is a major issue with respect to image quality reduction. These reflected regions appearing in image are called as specular reflections (SR). This problem is common in all types of images and it disturbs the image interpretation. Thus, the removal of SR pixels is one of the most important pre-processing steps for accurate image analysis. Several techniques are suggested in the literature to address this issue. The paper reports an in-depth review of various categories and issues of SR detection and the probable solution to overcome it. Experimental analysis proves that Kittler minimum error threshold selection method can be applied on input image as a preprocessing method for SR detection and analysis. Increase in Jaccard Index (JC) justifies the performance of proposed solution.