雾霾天气条件下偏振遥感图像复原研究

Research on Image Restoration by Polarized Remote Sensing Through Haze

针对雾霾天气造成图像质量退化引起的图像对比度降低问题,根据目标场景与大气气溶胶粒子散射光偏振特性的差异,开展雾霾天气目标偏振探测图像复原研究。通过实验获取雾霾天气条件下不同偏振旋转方位角度图像,提出图像复原算法进而提高图像对比度,抑制雾霾介质造成图像退化的影响,提高光学成像系统的成像质量。结果表明,此算法能有效提高图像对比度和图像信息熵值,恢复目标,同时获取了目标景物的深度信息。本文研究为雾霾天气下目标探测与识别提供了有效的技术途径。

To overcome the problem of a degraded image due to the reduction of contrast under hazy weather, according to the differing polarization characteristics between the target scene and atmos- pheric aerosol, we researched image restoration for target recognition based on polarized detection through haze. To obtain different polarized orientation images by experiment during hazy weather, we proposed an image restoration algorithm to enhance the image contrast, to remove degradation from haze on images. Experimental results show that this algorithm can effectively enhance image contrast and entropy as well as recover the target. As a by product, the method yields a range map of the scene. This research provides an effective method for the target recognition under hazy weather conditions.