The System of Shadow Detection and Removal Using Object Oriented Technique

From satellite like Quick Bird, IKONOS etc., very high resolution satellite images are available. There are usually affected with shadows. These shadows can reduce information from the images. In this paper, for the detection of shadow, an object oriented shadow detection of shadow an object oriented shadow detection method is used. In this technique, using segmentation process, each object in the input image is extracted. After that suspected shadows will be detected by comparing grayscale average of each object in segmented image and threshold obtained from original image. False shadows are eliminated by comparing grayscale average in different image bands. Then shadow removal will be done using boundary extraction and IOOPL matching. Homogeneous section will be attained through IOOPL matching. This is effective method for the detection and removal of shadow.