| 总结: | The visibility of images is seriously degraded by hazy weather, which will affect the tracking and recognition of targets. The severe and recurring haze in Southeast Asia is not only a humanitarian disaster but also a destructive force in optical remote sensing. Thus, recovering the true scene form a hazy image is necessary and significant. This project aims at surveying current haze removal approaches in the free and commercial software for both photography and remote sensing image, as well as developing and implementing a haze removal method for the imaging operation. Viability study for the removal of thin clouds may also be carries out, as haze and thin clouds may have many optical properties in common.
In this report, there are mainly two part. One is to survey and summarize the current haze removal techniques for image processing, followed by an experimental comparison of various image defogging algorithms/techniques. Another one is to implement the algorithm using MATLAB and do some further improvement.
|
|---|