光学 精密工程, 2017, 25 (8): 2212, 网络出版: 2017-10-16   

利用偏振信息的雾天图像全局最优重构

Global optimized hazed image reconstruction based on polarization information
作者单位
合肥工业大学 计算机与信息学院,安徽 合肥 230009
摘要
针对雾、霾等强散射环境下相机拍摄图像严重退化的问题,提出了一种降低雾霾环境对图像质量影响的图像去雾重构方法。基于雾天偏振成像模型,分别估计图像每个区域的重构参数,获取全局最优的重构参数。利用偏振滤波的方法估计雾天大气散射光的偏振度分布,利用自适应亮通道方法计算无穷远处大气散射光的强度分布,从而重构出去雾图像。最后,利用偏振度的纹理信息对重构图像进行增强。该方法考虑了图像中不同位置大气杂散光参数的不一致性,对图像每一区域的重构参数分别运算, 从而获得全局最优的重构参数图像。该方法还不要求图像必须包含天空区域,并且具有对灰度图像的处理效果较好的优点。
Abstract
Aimed at the serious degradation problem of camera images under strong scattering environments of fog, haze and so on, a reconstruction method of image defogging was proposed to reduce the influence of fog and haze environment on the image quality. On the basis of the polarization imaging model, the global optimized reconstruction parameters were obtained by estimating every pixel's parameters. Then, polarization filtering method was used to estimate the equivalent degree of polarization image of atmospheric scattering light and the auto adjusted light channel prior method was taken to calculate the distribution of atmospheric scattering light intensity. The global optimal reconstruction parameter image after defogging was obtained by the steps mentioned above. Finally, the texture information of polarization degree was used to enhance the image after defogging. The method pays attention in the inconsistencies of stray light parameters in different positions in images, and calculates every pixel's parameters respectively, so it captures the global optimal reconstruction parameter images. Moreover, the method does not require its images to contain the sky area, and has an advantage for gray image processing.

高隽, 毕冉, 赵录建, 范之国. 利用偏振信息的雾天图像全局最优重构[J]. 光学 精密工程, 2017, 25(8): 2212. GAO Jun, BI Ran, ZHAO Lu-jian, FAN Zhi-guo. Global optimized hazed image reconstruction based on polarization information[J]. Optics and Precision Engineering, 2017, 25(8): 2212.

本文已被 5 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!