利用偏振信息的雾天图像全局最优重构
发布时间：2019-03-27 15:17
【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]:In order to solve the problem of serious degradation of camera images in fog and haze, a method of image de-fogging and reconstruction is proposed to reduce the influence of haze environment on image quality. Based on the fog polarization imaging model, the reconstruction parameters of each region of the image are estimated, and the global optimal reconstruction parameters are obtained. Polarization filtering method is used to estimate the polarization distribution of scattered light in fog sky, and adaptive bright channel method is used to calculate the intensity distribution of scattered light in infinite distance, so as to reconstruct the fog image. Finally, the texture information of polarization degree is used to enhance the reconstructed image. The method takes account of the inconsistency of atmospheric stray light parameters at different positions in the image, and calculates the reconstruction parameters of each region of the image separately, so as to obtain the globally optimal reconstructed parameter images. This method does not require that the image must contain sky area, and it has the advantage of better processing effect for gray-scale image.
【作者单位】： 合肥工业大学计算机与信息学院;
【基金】：国家自然科学基金资助项目(No.61571177,No.61571175)
【分类号】：TP391.41
本文编号：2448301
[Abstract]:In order to solve the problem of serious degradation of camera images in fog and haze, a method of image de-fogging and reconstruction is proposed to reduce the influence of haze environment on image quality. Based on the fog polarization imaging model, the reconstruction parameters of each region of the image are estimated, and the global optimal reconstruction parameters are obtained. Polarization filtering method is used to estimate the polarization distribution of scattered light in fog sky, and adaptive bright channel method is used to calculate the intensity distribution of scattered light in infinite distance, so as to reconstruct the fog image. Finally, the texture information of polarization degree is used to enhance the reconstructed image. The method takes account of the inconsistency of atmospheric stray light parameters at different positions in the image, and calculates the reconstruction parameters of each region of the image separately, so as to obtain the globally optimal reconstructed parameter images. This method does not require that the image must contain sky area, and it has the advantage of better processing effect for gray-scale image.
【作者单位】： 合肥工业大学计算机与信息学院;
【基金】：国家自然科学基金资助项目(No.61571177,No.61571175)
【分类号】：TP391.41
【相似文献】
相关硕士学位论文 前3条
1 夏秀山;气溶胶散射下目标偏振信息提取及应用研究[D];合肥工业大学;2014年
2 王亚慧;双CCD同时偏振成像系统及其偏振信息处理软件系统[D];北京理工大学;2016年
3 崔民杰;光谱与偏振信息融合处理系统硬件设计[D];南京理工大学;2016年
，本文编号：2448301
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2448301.html
