Local contrast preserving technique for the removal of thin cloud in aerial image

Abstract

This paper presents a local contrast preserving technique based on multi-scale retinex with color restoration (MSRCR) for the aerial image. We first propose a divided strategy of the aerial image through the estimation of the intensity distribution of width and height direction. The aerial image is then divided into multiple blocked images according to the divided strategy. Afterwards, one contrast factor for each blocked image is calculated and the multiple local optimal thresholds are obtained from the blocked image. As a result the thin cloud is removed and contrast is restored through these local optimal factors in the aerial image. Experimental results show that the improved technique can effectively distinguish details and greatly enhance the contrast in the aerial image processed.