一种基于幂指数拉伸的去雾算法

doi:10.16337/j.1004-9037.2022.01.005

首页 > 按月查看>2022年第1月 >62-72. DOI:10.16337/j.1004-9037.2022.01.005

一种基于幂指数拉伸的去雾算法
DOI:
                        10.16337/j.1004-9037.2022.01.005
                    
作者:
                        
                        
                    
作者单位:1.江苏科技大学机械工程学院，镇江212000;2.贵州风雷航空军械有限责任公司，安顺561000
作者简介:
通讯作者:
基金项目:国家重点研发计划（2018YFC0309100）。

Defogging Algorithm Based on Power Exponent Stretching

Author:

Affiliation:

1.College of Mechanical Engineering,Jiangsu University of Science and Technology,Zhenjiang 212000,China;2.Guizhou Fenglei Aviation Armament Co., Ltd.,Anshun 561000,China

Fund Project:

摘要

图/表

访问统计

参考文献

相似文献

引证文献

资源附件

摘要:

比较同一场景无雾和有雾时图像RGB（Red-green-blue）三通道和HSV（Hue-saturation-value）三通道的变化，提出一种基于幂指数拉伸的去雾算法。首先将图像从RGB变换到HSV空间，将饱和度分量和亮度分量分别作1~3的幂指数拉伸和调整，将拉伸变换后分量生成HSV图像再变换到RGB空间，生成增强后的去雾图像。以饱和度均值、亮度指标、信息熵和对比度作为去雾评价的指标，确定最优的拉伸幂指数组合。然后使用最优幂指数完成去雾处理，同时根据图像饱和变化的阈值或时间间隔长度决定是否重新寻找最优拉伸幂指数。最后使用Python软件，借助多进程编程实现本文去雾算法。当图像分辨率为400像素×300像素时，树莓派上运行时幂指数参数寻优用时为5.077~6.160 s，单帧图像去雾用时第1帧时间长为0.308 s，其余时间为0.077~0.168 s，结果验证了本文算法的实时性。

Abstract:

After comparing three channels of RGB（Red-green-blue） and three channels of HSV（Hue-saturation-value） in the same scene between clear and fog pictures， a haze removal algorithm based on power exponent stretching is proposed. Firstly， the image is transformed from RGB to HSV space. Then the saturation component and the brightness component are exponentially stretched with power of 1—3，and then they are both adjusted to their suitable range. After stretching transformation of saturation and brightness， the image is transformed from HSV to RGB space to generate enhanced defogging images. Taking the mean value of saturation， brightness index， information entropy and contrast as defog evaluation indexes， the optimal stretching power index combination is determined. The optimal power index combination is used to complete the defogging process. At the same time， it is decided whether to find the optimal power index again according to the change of image average saturation or the length of time interval. Finally， the fog removal algorithm is implemented by multi-process programming with the Python software. When the image resolution is 400 pixel×300 pixel， it takes 5.077—6.160 s to optimize the power index parameters on the raspberry PI. For one frame defogging， the first frame takes longer time of 0.308 s. The other frames take 0.077—0.168 s to removal haze for a single frame.

表 1 高楼场景不同点的RGB和HSV空间分量值Table 1 Values of RGB and HSV space components of different points in the building scene

图1 操场场景的无雾和有雾图像Fig.1 Fog-free and foggy images of the playground scene

图2 无雾和有雾图像R、G、B三分量的比较Fig.2 Comparison of three components R,G and B of fog-free and foggy images

图3 无雾和有雾图像H、S、V三分量比较Fig.3 Comparison of three components H,S and V of fog-free and foggy images

图4 无雾图像和有雾图像的H、S、V分量效果比较Fig.4 Effect comparison of H，S，V components of fog-free and foggy images

图5 高楼场景的无雾图像和浓雾图像Fig.5 Fog-free image and dense fog image of the building scene

图6 幂指数拉伸增强去雾算法流程图Fig.6 Flow chart of power exponential stretching based dehazing enhancment

图7 图1（b）幂指数拉伸后去雾效果图Fig.7 Dehazing image of Fig.1(b) after power exponent stretching

图8 3种去雾方法效果比较Fig.8 Dehazing effect comparison of three methods

图9 本文算法对于不同场景去雾效果比较（左边为有雾图像，右边为去雾图像）Fig.9 Comparison of defogging effects in different scenes (fogging images on the left, defogging images on the right)