Effect of shadow removal by gamma correction in SMQT algorithm in environmental application

Environmental data can be achieved from remote sensing. Shadows are main problems in achieving environmental data by remote sensing. In this research, an algorithm is proposed for preparing environmental data from shadow-covered objects. The proposed algorithm provides some information about digital number (DN) of shadow-covered objects by reducing the shadow effect. The successive mean quantization transform (SMQT) algorithm, gamma correction and combination of images form the basis of the proposed algorithm. SMQT algorithm eliminates gain and bias features by indicating the data structure. The data structure remains unchanged, and the differences between terrains are highlighted by compressing the dynamic range and stretching the histogram of the image in parts with different terrains of same DN. The output of the SMQT algorithm is a gray image. A new image is obtained by combining this image with the original image through HSV color space. In the second step, gamma correction is applied to the entire image according to brightness and contrast. But the gamma correction rate is not the same in all parts of an image. As a result, gamma correction should be done locally. However, local gamma correction and the use of a kernel of a specific dimension increase computation time. In addition, if there is a noise in the image, it will cause a significant deviation in the correction. To solve this problem, the output of SMQT algorithm is combined with the image obtained from gamma correction. Shadows in the input image cause the image histogram to be compressed in areas near zero. After performing the process in the study, the histogram was stretched and the maximum histogram reached from 190 to 250.