Research on Sparse Code Shrinkage denoise in Underwater 3D Laser Scanning Images

Abstract

Noise removal is the pre-stage and essential step for many image processing tasks. Sometimes it is hard to have correct exposure for camera in underwater environment. In general, we need to have long exposure time to collect the weak light signal which increase the background electrical noise and higher than the signal. Underwater laser scanning is such special case with white laser stripe and dark background images. Hence, noise including which dots, lumps have heavy impact to our laser image. This paper discuss underwater laser scanning image denoise and presents the method based on Sparse code shrinkage algorithm to reduce such noise effect in final reconstructed 3D model. The comparative study using natural and real 3D underwater scanning images is presented on the experimental section.