An auto-threshold control for isolating false shadow from high-resolution images

Abstract

Digital image processing helps to extract the useful information from images for further investigations and analysis. Pre-processing of images, such as removal of noise and shadows, improves the quality of the high-resolution images for image recognition and tracking. In this paper, an optimal filter is recommended from its performance metrics for reducing speckle noise from the high-resolution images. The objective of the paper is to isolate the deceptive shadow from the original shadow, which is dark in color but is falsely treated as a shadow. Purging the false shadow is done by exploring the appropriate threshold value of a shadow region. Experiments show that our proposed method with auto-threshold control helps to isolate the deceptive shadow.