Boosting images quality of environment remote sensing based on Enhanced Contrast Stretching Algorithm

Yuan An,Zhaohui Cai,Yingrui Ma,Chun Liu,Hongbo Li

DOI: https://doi.org/10.1088/1755-1315/675/1/012178

2021-02-01

IOP Conference Series: Earth and Environmental Science

Abstract:Abstract Remote sensing, especially the remote satellite images monitoring, plays an important role in environment research domain. They are widely used for environment statistics, changing patters discovery and prediction tasks. However, the raw remote sensing images are normally faced with noise problems and hard to do visual interpretation analysis, which also hurts the following environmental analysis, data mining and prediction tasks. To solve this challenge, we propose an Enhanced Contrast Stretching Algorithm (ECSA) based on Landsat ETM + images. The proposed algorithm converts the distance between the gray value of each pixel and the gray mean value of each single-band image into approximately normal distribution, and constructs a new gray scale transformation function on this basis to enhance the gray stretch ratio of each pixel, thereby realizing images enhancement. The experimental results on real-world data show that ECSA can effectively improve the images quality in environmental monitoring images, and acquire better color naturalness and detailed characteristics of environmental research targets. As a pre-processing images enhancement procedure, our proposed method could not only boost the images quality, but also, be conducive to do the visual interpretation of remote sensing images in environmental monitoring area, and can effectively improve the accuracy of subsequent research tasks.