交通图像大气加雾模型设计与评价

doi:10.19562/j.chinasae.qcgc.2023.05.008

汽车工程 ›› 2023, Vol. 45 ›› Issue (5): 786-795.doi: 10.19562/j.chinasae.qcgc.2023.05.008

所属专题：智能网联汽车技术专题-感知&HMI&测评2023年

交通图像大气加雾模型设计与评价

黄鹤^1,²,李战一^1,²,杨澜³(),王会峰¹,高涛³,陈婷³

^1.长安大学电子与控制工程学院，西安　710064
^2.西安市智慧高速公路信息融合与控制重点实验室，西安　710064
^3.长安大学信息工程学院，西安　710064

收稿日期:2022-09-16 修回日期:2022-12-09 出版日期:2023-05-25 发布日期:2023-05-26
通讯作者: 杨澜 E-mail:lanyang@chd.edu.cn
基金资助:
国家重点研发计划项目(2021YFB2501200);国家自然科学基金面上项目(52172379);陕西省重点研发计划项目(2022GY-300);西安市智慧高速公路信息融合与控制重点实验室（长安大学）开放基金项目(300102323502);中央高校基本科研业务费专项资金重点科研平台建设计划水平提升项目(300102323501)

Design and Evaluation of Atmospheric Fogging Model for Traffic Image

He Huang^1,²,Zhanyi Li^1,²,Lan Yang³(),Huifeng Wang¹,Tao Gao³,Ting Chen³

^1.School of Electronics and Control Engineering，Chang'an University，Xi’an 　710064
^2.Xi 'an Key Laboratory of Intelligent Expressway Information Fusion and Control，Xi’an 　710064
^3.School of Information Engineering，Chang'an University，Xi’an 　710064

Received:2022-09-16 Revised:2022-12-09 Online:2023-05-25 Published:2023-05-26
Contact: Lan Yang E-mail:lanyang@chd.edu.cn

摘要/Abstract

摘要：

针对数字孪生过程中，交通雾霾图像的采集受天气限制，数据库获取困难导致样本不足等问题，提出了一种新的大气加雾模型，并用于扩展不同浓度的交通雾霾图像数据库。首先，结合暗特征原理，求解大气光值，并提出了一种基于区域方差的大气光补偿方法来获取大气光估计；其次，利用颜色衰减先验估计场景深度，求解初始透射率；然后构建了图像大气加雾模型，将计算的大气光估计与大气加雾透射率代入模型，并利用雾霾系数调整加雾浓度；最后设计了多组交通视频加雾实验并进行评价。实验结果表明，提出的算法能随着预置雾霾系数增大，使得图像主观上明显趋于模糊，客观指标随之逐步发生变化，图像降质规律与真实的含雾场景基本一致，可用于扩充雾霾数据集，具有很好的有效性和实用性。利用不同去雾算法评价对比加雾图像可知，复原图像效果与针对实际图像的去雾效果基本无异，进一步反向验证了加雾模型的有效性。

关键词: 图像处理, 图像加雾, 数字孪生, 大气加雾模型

Abstract:

For the problems of weather restriction and insufficient samples due to the difficulty of database acquisition for the collection of traffic haze images in the process of digital twinning, a new atmospheric fogging model is proposed to expand the database of traffic haze images with different concentrations. Firstly, the atmospheric light value is calculated based on the principle of dark feature, and the estimation of atmospheric light is obtained by using the atmospheric light compensation method based on variance fluctuation. Secondly, the color attenuation prior is used to estimate the scene depth, and to solve the initial transmittance. Then, an image atmospheric fogging model is constructed, and the calculated atmospheric light estimation and atmospheric fogging transmission are substituted into the model, and the fogging density is adjusted by the haze coefficient. Finally, several traffic video fogging experiments are designed and evaluated. Experimental results show that with the increase of preset haze coefficient, the proposed algorithm can make the image tend to blur subjectively, and the objective index gradually change accordingly. The image quality degradation law is basically consistent with the real scene containing fog, which can be used to expand haze data set, with good effectiveness and practicability. By evaluating and comparing the fogging images with different defogging algorithms, it can be seen that the effect of the restored image is basically the same as the defogging effect of the actual image, further verifying the effectiveness of the fogging model.

Key words: image processing, image fogging, digital twin, fogging model

黄鹤,李战一,杨澜,王会峰,高涛,陈婷. 交通图像大气加雾模型设计与评价[J]. 汽车工程, 2023, 45(5): 786-795.

He Huang,Zhanyi Li,Lan Yang,Huifeng Wang,Tao Gao,Ting Chen. Design and Evaluation of Atmospheric Fogging Model for Traffic Image[J]. Automotive Engineering, 2023, 45(5): 786-795.

图/表 25

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

图12

图13

图14

图15

图16

表1

图17

图18

图19

图20

图21

图22

图23

图24

参考文献 21

1	WANG K， WANG H， LI Y， et al.Quantitative performance evaluation for dehazing algorithms on synthetic outdoor hazy images［J］. IEEE Access， 2018， 6：20481-20496. doi： 10.1109/ACCESS.2018.2822775.
2	TAN R T. Visibility in bad weather from a single image［C］. Proceedings of the 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition （CVPR）， 2008：1-8.
3	OAKLEY J P， SATHERLEY B L. Improving image quality in poor visibility conditions using a physical model for contrast degradation［J］. IEEE Transactions on Image Processing， 1998， 7（2）：167-179.
4	CAI B， XU X， JIA K， et al. DehazeNet： an end-to-end system for single image haze removal ［J］. IEEE Transactions on Image Processing， 2016， 25（11）： 5187-5198.
5	LI B， PENG X， WANG Z， et al.AOD-Net： all-in-one dehazing network［C］.2017 IEEE International Conference on Computer Vision （ICCV）， 2017： 4780-4788. doi： 10.1109/ICCV.2017.511.
6	ZHANG H， PATEL V M.Densely connected pyramid dehazing network［C］.2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition， 2018：3194-3203. doi： 10.1109/CVPR.2018.00337.
7	HE K， SUN J， TANG X. Single image haze removal using dark channel prior ［J］. IEEE Transactions on Pattern Analysis & Machine Intelligence， 2011， 33（12）： 2341-2353.
8	HE K， SUN J， TANG X. Guided image filtering［C］. Proceedings of the European Conference on Computer Vision（ECCV）， 2010：1-14.
9	HE K， SUN J. Fast guided filter［R］. Computer Science， 2015.
10	ZHU Q， MAI J， SHAO L. A fast single image haze removal algorithm using color attenuation prior ［J］. IEEE Transactions on Image Processing， 2015， 24（11）： 3522-3533.
11	BERMAN D， TREIBITZ T， AVIDAN S. Non-local image dehazing［C］. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition （CVPR），2016： 1674-1682.
12	BERMAN D， TREIBITZ T， AVIDAN S. Air-light estimation using haze-lines［C］.2017 IEEE International Conference on Computational Photography （ICCP）. IEEE， 2017.
13	肖进胜，周景龙，雷俊锋，等.面向图像场景转换的改进型生成对抗网络［J］.软件学报，2021，32（9）：2755-2768.
	XIAO Jinsheng， ZHOU Jinglong， LEI Junfeng， et al. Improved generative adversarial network for image scene transformation［J］. Journal of Software， 2021，32（9）：2755-2768.
14	WANG K， WANG H， LI Y， et al. Quantitative performance evaluation for dehazing algorithms on synthetic outdoor hazy images［J］. IEEE Access， 2018， PP：1-1.
15	MCCARTNEY E J， et al. Optics of the atmosphere： scattering by molecules and particles［J］. Physics Today， 1977，30（5）：76.
16	NARASIMHAN S G，NAYAR S K. Contrast restoration of weather degraded images［J］. IEEE Trans. Pattern Anal. Mach. Learn.， 2003，25（6）：713-724.
17	NAYAR S K， NARASIMHAN S G. Vision in bad weather［C］. Proceedings of the Seventh IEEE International Conference on Computer Vision（ICCV），2002： 1-9.
18	CHOI L K， YOU J， BOVIK A C. Referenceless prediction of perceptual fog density and perceptual image defogging ［J］. IEEE Transactions on Image Processing， 2015， 24（11）： 3888-3901.
19	HUANG H， SONG J， GUO L， et al. Haze removal method based on a variation function and colour attenuation prior for UAV remote-sensing images ［J］. Journal of Modern Optics， 2019， 66（2）： 1-14.
20	黄鹤，胡凯益，宋京，等. 雾霾线求解透射率的二次优化方法［J］. 西安交通大学学报， 2021， 55（8）：130-138.
	HUANG He， HU Kaiyi， SONG Jing， et al. A twice optimization method for solving transmittance with haze-lines［J］. Journal of Xi’an Jiaotong University， 2021， 55（8）：130-138.
21	GALDRAN A， VAZQUEZ-CORRAL J， PARDO D， et al. Fusion-based variational image dehazing［J］. IEEE Signal Processing Letters， 2017， 24（2）：151-155.

指标	原图	α=0.1	α=0.3	α=0.5	α=0.7	α=0.9
信息熵	14.968	14.177	13.437	12.662	11.884	11.113
平均梯度	7.716 4	7.076 2	5.776 2	4.704 3	3.820 5	3.087 7
FADE	0.487 2	0.805 4	1.352 7	1.948 1	2.570 9	3.217 6
PSNR		24.859	24.116	24.094	24.091	24.090
SSIM		0.828 6	0.657 5	0.520 0	0.410 0	0.321 0
模糊系数		0.958 0	0.841 5	0.738 3	0.652 8	0.581 9
SIFT算子	2 945	2 835	2 403	2 007	1 649	1 315
OBR算子	5 419	2 709	2 287	1 860	1 446	1 097

[1]	姚宇捷,彭育辉,陈泽辉,何维堃,吴庆,黄炜,陈文强. 支持抗光照目标检测的改进YOLO算法[J]. 汽车工程, 2023, 45(5): 777-785.
[2]	吴桐,王宇宁,关艺搏,田韶鹏. 基于改进门控上下文聚合网络的雾天实时行人检测方法[J]. 汽车工程, 2023, 45(5): 796-806.
[3]	赵树廉,来飞,李克强,陈涛,孟璋劼,唐逸超,吴思宇,田浩东. 基于数字孪生技术的智能汽车测试方法研究[J]. 汽车工程, 2023, 45(1): 42-51.

交通图像大气加雾模型设计与评价

Design and Evaluation of Atmospheric Fogging Model for Traffic Image

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 25

参考文献 21

相关文章 3

Metrics

本文评价

推荐阅读 10