Thin cloud removal for satellite images using normalized intensity mapping

Clouds inhibit the features of images such as color and brightness in the image regions. The light diffusion, scattering and attenuation of clouds create blur effect or affect the contrast of the images. The interference of cloud over images degrades the quality of satellite images and tends to suppress the essential information in those images. Hence, the removal of clouds greatly help in the efficient processing of satellite images. The proposed cloud removal method, Normalized Intensity Mapping (NIM) is confirmed to be effective for cloud removal. This technique was experimented on the cloud covered real-time images and was found to outperform its competitive techniques. The quality of NIMCR is assessed quantitatively by Peak Signal to Noise Ratio (PSNR) and qualitatively by Human Visual Perception (HVP). This technique due to its applicability in real-time and satellite images, qualifies itself as a valuable addition in the array of image pre-processing techniques.