Performance Evaluation of Quality Enhancement Methods of Remote Sensing Images by Object Oriented Shadow Detection and Removal

Abstract: Shadow will occur by sunlight or any light sources. In the area of machine vision, shadows occur frequently in a wide variety of scenes. In many cases, this is undesirable due to the fact that they often lead to the result of irretrievable processing failures. In this paper, we introduce a novel shadow detection and removal technique that produces a shadow-free scene. In this method, shadow features are taken into consideration during image segmentation, after that an SVM based classifier is used for the accuracy and then, according to the statistical features of the images, suspected shadows are extracted. Dark objects may be included in the suspected shadows, so more accurate shadow detection results are needed to eliminate these dark objects. For shadow removal, inner–outer outline profile line (IOOPL) matching is used. First, the IOOPLs are obtained with respect to the boundary lines of shadows. Shadow removal is then performed according to the homogeneous sections attained through IOOPL similarity matching. Finally, using the homogeneous sections, the relative radiation calibration parameters between the shadow and non- shadow regions are obtained, and shadow removal is performed. Also another approach based on YCbCr colour space is used for the shadow detection process and the performance evaluation of the both methods is carried out. The new methods can accurately detect shadows from urban high-resolution remote sensing images and can effectively restore shadows.

Keywords: Object-oriented, remote sensing images (RSI), change detection, support vector machine (SVM) classifier, inner–outer outline profile line (IOOPL), relative radiometric correction, shadow detection, shadow removal, mean, standard deviation.