一种改进的透射率分布估计的夜间图像去雾算法

薛楠; 严利民

一种改进的透射率分布估计的夜间图像去雾算法

薛楠^1,,
严利民^{1, 2, ,}

1.
上海大学机电工程与自动化学院, 上海 200444
2.
上海大学微电子研究与开发中心, 上海 200444

详细信息

作者简介:
薛楠（2001-），男，本科生，研究领域为图像处理及计算机视觉。E-mail：wilson_xue@outlook.com

通讯作者:
严利民（1971-），男，博士，副教授，研究领域为集成电路设计及系统集成、新型显示技术和计算机视觉。E-mail：yanlm@shu.edu.cn

中图分类号: TP391
计量
- 文章访问数: 146
- HTML全文浏览量: 31
- PDF下载量: 34
- 被引次数: 0
出版历程
- 收稿日期: 2022-01-02
- 修回日期: 2022-03-24
- 刊出日期: 2022-10-20

Nighttime Image Dehazing Algorithm Based on Improved Transmittance Distribution Estimation

XUE Nan^1
,,
YAN Limin^{1, 2
, ,}

1.
School of Mechatronic Engineering and Automation, Shanghai University, Shanghai 200444, China
2.
Microelectronics R & D Center, Shanghai University, Shanghai 200444, China

摘要

摘要: 针对基于暗通道先验理论（dark channel prior, DCP）的去雾算法在处理夜间有雾图像时细节信息缺失、光源区域的纹理受损严重的问题，本文提出了一种改进的透射率分布估计的夜间图像去雾算法。通过引入暗态点光源模型、暗通道可信度权值因子和伪去雾图像，结合夜间图像成像模型，获取改进的透射率分布，对夜间降质图像进行去雾处理。实验结果表明，经本文算法处理后的图像在纹理细节上损失小、图像清晰度高，图像明暗对比度得到较好的拉伸，可以实现夜间有雾图像的有效去雾。
- 夜间图像去雾 /
- 暗通道先验 /
- 透射率分布估计
Abstract: This paper presents an improved transmittance distribution estimation algorithm for nighttime image dehazing to solve lack of detailed information and serious damage to the texture of light source areas when the dark channel prior dehazing algorithm processes foggy images at night. An improved transmittance distribution was obtained by introducing a dark state point light source model, a dark channel credibility weight factor, and a pseudo dehazing image, combined with a nighttime image imaging model, and the dehazed image at night was dehazed. The experimental results showed that the image processed by using the proposed algorithm had little loss in texture details and high image definition, and the contrast between the light and dark of the image was better stretched, which effectively dehazed a foggy image at night.
- nighttime image dehazing /
- dark channel prior /
- transmission distribution estimation

HTML全文

图 1 采用暗态点光源模型的夜间图像去雾算法流程

Figure 1. Process of nighttime image dehazing algorithm using dark point light source model

下载: 全尺寸图片幻灯片

图 2 暗通道可信度权值因子获取流程

Figure 2. Dark channel credibility weight factor acquisition process

下载: 全尺寸图片幻灯片

图 3 扇形模型几何表示

Figure 3. Sector model geometric representation

下载: 全尺寸图片幻灯片

图 4 本文算法的去雾结果对比

Figure 4. Comparison of dehazing results of the algorithm in this paper

下载: 全尺寸图片幻灯片

表 1 图像信息熵统计结果

Table 1. Image information entropy

Comparative object	Original image	Enhanced image	Algorithm 1	Algorithm 2	Algorithm 3	Method of this paper
Fig. 1	6.87	7.15	7.28	7.46	9.04	9.77
Fig. 2	6.74	7.03	7.26	7.2	7.21	7.61
Fig. 3	7.14	7.95	7.29	8.44	11.17	13.81
Fig. 4	6.57	7.28	7.68	8.06	8.98	9.74

下载: 导出CSV

表 2 图像平均梯度统计结果

Table 2. Image average gradient

Comparative object	Original image	Enhanced image	Algorithm 1	Algorithm 2	Algorithm 3	Method of this paper
Fig. 1	4.21	6.59	8.13	10.51	7.19	13.61
Fig. 2	1.56	2.56	3.56	4.42	7.83	8.04
Fig. 3	9.77	13.01	14.41	17.79	19.05	22.11
Fig. 4	5.53	8.92	9.09	14.22	11.95	17.16

下载: 导出CSV

表 3 图像对比度统计结果

Table 3. Image contrast

Comparative object	Original image	Enhanced image	Algorithm 1	Algorithm 2	Algorithm 3	Method of this paper
Fig. 1	61.05	202.34	247.98	365.75	446.84	558.29
Fig. 2	14.23	23.58	28.97	50.72	47.31	51.79
Fig. 3	229.85	493.12	581.56	707.65	943.99	976.51
Fig. 4	82.88	196.79	231.62	320.78	337.73	441.38

下载: 导出CSV

参考文献(10)

[1]	HE K, SUN J, TANG X. Single image haze removal using dark channel prior[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 33(12): 2341-2353.
[2]	ZHANG J, CAO Y, WANG Z F. Nighttime haze removal based on a new imaging model[C]//IEEE International Conference on Image Processing, 2014: 4557-4561.
[3]	PEI S C, LEE T Y. Nighttime haze removal using color transfer pre-processing and dark channel prior[C]// Image Processing (ICIP), 19th IEEE International Conference, 2012: 957-960.
[4]	陈志恒, 严利民, 张竞阳. 采用自适应全局亮度补偿的夜间去雾算法[J]. 红外技术, 2021, 43(10): 954-959. https://www.cnki.com.cn/Article/CJFDTOTAL-HWJS202110007.htm CHEN Zhiheng, YAN Limin, ZHANG Jingyang. Nighttime dehazing algorithm with adaptive global brightness compensation[J]. Infrared Technology, 2021, 43(10): 954-959. https://www.cnki.com.cn/Article/CJFDTOTAL-HWJS202110007.htm
[5]	皮燕燕, 王远明, 王兴岭. 基于图像分层与暗通道先验的夜间图像去雾算法[C]//惯性技术发展动态发展方向研讨会文集, 2018: 132-135. PI Yanyan, WANG Yuanming, WANG Xingling. Night-time image dehazing algorithm based on image stratification and dark channel prior[C]// Proceedings of the Symposium on the Dynamic Development Direction of Inertial Technology, 2018: 132-135.
[6]	王柳哲. 基于多光源模型与暗通道先验的夜间图像去雾[D]. 北京: 北京交通大学, 2018. WANG Liuzhe. Night Image Defogging Based on Multiple Light Source Models and Dark Channel Priors[D]. Beijing: Beijing Jiaotong University, 2018.
[7]	张竞阳, 严利民, 陈志恒. 采用暗态点光源模型的夜间去雾算法[J]. 红外技术, 2021, 43(8): 798-803. https://www.cnki.com.cn/Article/CJFDTOTAL-HWJS202108015.htm ZHANG Jingyang, YAN Limin, CHEN Zhiheng. Nighttime fog removal using the dark point light source model[J]. Infrared Technology, 2021, 43(8): 798-803. https://www.cnki.com.cn/Article/CJFDTOTAL-HWJS202108015.htm
[8]	ZHANG J, CAO Y, FANG S, et al. Fast haze removal for nighttime image using maximum reflectance prior[C]//IEEE Conference on Computer Vision and Pattern Recognition, 2017: 7016-7024.
[9]	YIN Hui, GONG Yuanhao, QIU Guoping. Side window guided filtering[J]. Signal Processing, 2019, 165(C): 315-330.
[10]	高强, 胡辽林, 陈鑫. 基于暗通道补偿与大气光值改进的图像去雾方法[J]. 激光与光电子学进展, 2020, 57(6): 150-156. https://www.cnki.com.cn/Article/CJFDTOTAL-JGDJ202006015.htm GAO Qiang, HU Liaolin, CHEN Xin. Image dehazing method based on dark channel compensation and improvement of atmospheric light value[J]. Laser & Optoelectronics Progress, 2020, 57(6): 150-156. https://www.cnki.com.cn/Article/CJFDTOTAL-JGDJ202006015.htm