雾霾污染下基于混沌性的图像去雾算法改进-《郑州大学学报(工学版)》

文章信息/Info

Title:: Image Remove the Fogfog Algorithm Based on Improvement Chaotic under Haze Pollution

摘要:: 传统基于边缘保持滤波的单幅图像快速去雾算法,在雾霾污染大气粒子散射作用下,图像会受到雾化背景的干扰,出现浓雾噪点,图像结构信息复原效果较差.提出一种基于混沌性的加权滤波图像快速去雾算法,通过分析雾天雾化背景干扰下的图像信息,获取图像数据的混沌特性,并对雾化图像进行最小颜色分量估计.在此基础上,结合带雾图像暗原色模型、时域和频域特征分量模型,构建出自适应加权滤波模型,完成基于混沌性的加权滤波图像快速去雾算法的改进设计.实验结果表明,采用该算法能避免雾化图像中间区域的颜色失真,降低雾化背景干扰,减少浓雾噪点,使远景的混沌特征得到合理的保留,在图像质量和运算性能方面都具有优越性.

Abstract:: The traditional edge preserving filtering based single image fast defogging algorithm, under the influence of atmospheric particle scattering caused by haze pollution, the image will be affected by the fog background, resulting in dense fog noise and poor image structure information restoration. A weighted filtering image fast defogging algorithm based on chaos is proposed, which obtains the chaotic characteristics of the image data by analyzing the image information under the fog background interference, And the minimum color component estimation is performed on the atomized image. Based on this, a self adaptive weighted filtering model is constructed by combining the dark primary color model of the fogged image, the time-domain and frequency-domain feature component models, and an improved design of a chaotic weighted filtering image fast defogging algorithm is completed. The experimental results show that using this algorithm can avoid color distortion in the middle area of the atomized image, reduce atomization background interference, and reduce dense fog noise, Reasonably preserving the chaotic features of the distant view has advantages in image quality and computational performance