基于预检机制的偏振图像去烟研究

大气与环境光学学报 ›› 2023, Vol. 18 ›› Issue (2): 108-118.

基于预检机制的偏振图像去烟研究

阎庆 1, 叶孟孟 1,2*, 张晶晶 1,2,3, 刘晓 3, 年福东 1,4, 李腾 1,2

1 计算智能与信号处理教育部重点实验室 (安徽大学), 安徽合肥 230601; 2 偏振光成像探测技术安徽省重点实验室, 安徽合肥 230031; 3 中国科学院通用光学定标与表征技术重点实验室, 安徽合肥 230031; 4 合肥学院先进制造工程学院, 安徽合肥 230031

收稿日期:2021-06-25 修回日期:2021-09-09 出版日期:2023-03-28 发布日期:2023-04-18
通讯作者: 叶孟孟 E-mail:1156761097@qq.com
作者简介:阎庆( 1978- ), 女, 安徽合肥人, 博士, 副教授, 硕士生导师, 主要研究方向为模式识别、图像处理和深度学习。 E-mail: rubby_yan5996@sina.com
基金资助:
中国科学院通用光学定标与表征技术重点实验室开放研究基金, 偏振光成像探测技术安徽省重点实验室开放基金, 国家自然科学基金青年科学基金 (61902104), 安徽省自然科学基金 (2008085QF295), 安徽高校自然科学研究项目 (KJ2020A0651)

Polarization image smoke removal based on precheck mechanism

YAN Qing 1, YE Mengmeng 1,2*, ZHANG Jingjing 1,2,3, LIU Xiao 3, NIAN Fudong 1,4, LI Teng 1,2

1 Key Laboratory of Computational Intelligence and Signal Processing (Anhui University), Ministry of Education, Hefei, Anhui 230601, China; 2 Anhui Key Laboratory of Polarized Light Imaging Detection Technology, Hefei 230031, China; 3 Key Laboratory of Optical Calibration and Characterization, Chinese Academy of Sciences, Hefei 230031, China; 4 School of Advanced Manufacturing Engineering, Hefei University, Hefei 230031, China

Received:2021-06-25 Revised:2021-09-09 Published:2023-03-28 Online:2023-04-18

摘要/Abstract

摘要： 烟的存在会导致图像目标信息的损减或丢失。针对烟在场景中具有局域性，提出了基于目标检测Yolov3 算法的去烟预检机制，即在去烟流程中增加预检机制实现对烟图定向去烟，提升去烟效率和避免对无烟区的影响。不同于现有针对可见光图像的基于深度学习去雾方法，该方法将四幅偏振态图像作为网络输入，并利用多尺度注意力对抗网络提取烟区目标的偏振态特征信息，从而缓解失真现象以及丰富去烟后目标的结构和细节信息。在真实数据集上的定性与定量实验结果表明，本文提出的算法有效提升了偏振图像的去烟效果和去烟效率。

关键词: 图像除烟, 卷积网络, 偏振态图像, 多尺度, 注意力机制, 对抗网络

Abstract: The presence of smoke can cause the damage or loss of image target information. In view of the local nature of smoke in the scene, a smoke removal precheck mechanism based on the target detection Yolov3 algorithm is proposed in this work, that is, a precheck mechanism is added in the smoke removal process to realize the directional removal of smoke on the smoke image, improve the efficiency of smoke removal and avoid the impact of smoke on the non-smoking area. Different from the existing deep learningbased defogging methods for visible images, this method takes four polarization images as network input,and uses multi-scale attention adversarial network to extract the polarization information of the target in the smoke area, so as to alleviate distortion and enrich the structure and detail information of the target after smoke removal. Qualitative and quantitative experimental results on real data sets show that the proposed algorithm can effectively improve the smoke removal effectiveness and efficiency of polarized images.

Key words: image smoke removal, convolutional network, polarization image, multi-scale, attention mechanism, adversarial network

中图分类号:

阎庆, 叶孟孟, 张晶晶, 刘晓, 年福东, 李腾, . 基于预检机制的偏振图像去烟研究[J]. 大气与环境光学学报, 2023, 18(2): 108-118.

YAN Qing , YE Mengmeng , ZHANG Jingjing , LIU Xiao , NIAN Fudong , LI Teng , . Polarization image smoke removal based on precheck mechanism[J]. Journal of Atmospheric and Environmental Optics, 2023, 18(2): 108-118.

参考文献 21

[1]	He K M, Sun J, Tang X O. Single image haze removal using dark channel prior [J]. IEEE Transactions on Pattern Analysis
	and Machine Intelligence, 2011, 33(12): 2341-2353.
[2]	Jiang Y T, Sun C M, Zhao Y, et al. Image dehazing using adaptive bi-channel priors on superpixels [J]. Computer Vision and
	Image Understanding, 2017, 165: 17-32.
[3]	Zhao L J, Gao J, Bi R, et al. Polarization defogging method based on maximum and minimum intensity images [J]. Journal
	of Applied Optics, 2017, 38(3): 415-420.
	赵录建, 高隽, 毕冉, 等. 基于最大和最小光强图像的偏振去雾方法 [J]. 应用光学, 2017, 38(3): 415-420.
[4]	Gao J, Bi R, Zhao L J, et al. Global optimized hazed image reconstruction based on polarization information [J]. Optics and
	Precision Engineering, 2017, 25(8): 2212-2220.
	高隽, 毕冉, 赵录建, 等. 利用偏振信息的雾天图像全局最优重构 [J]. 光学精密工程, 2017, 25(8): 2212-2220.
[5]	Chen D D, He M M, Fan Q N, et al. Gated context aggregation network for image dehazing and deraining [C]. 2019 IEEE
	Winter Conference on Applications of Computer Vision (WACV). January 7-11, 2019, Waikoloa, HI, USA. IEEE, 2019:
13	75-1383.
[6]	Qin X, Wang Z L, Bai Y C, et al. FFA-net: Feature fusion attention network for single image dehazing [J]. Proceedings of the
	AAAI Conference on Artificial Intelligence, 2020, 34(7): 11908-11915.
[7]	Liao Y B. Polarized Light Science [M]. Beijing: Science Press, 2003: 59-62.
	廖延彪. 偏振光学 [M]. 北京: 科学出版社, 2003: 59-62.
[8]	Chen Y P, Fan H Q, Xu B, et al. Drop an octave: Reducing spatial redundancy in convolutional neural networks with octave
	convolution [C]. 2019 IEEE/CVF International Conference on Computer Vision (ICCV). October 27-November 2, 2019,
	Seoul, Korea (South). IEEE, 2020: 3434-3443.
[9]	Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need [C]. Advances in Neural Information Processing Systems.
20	17: 5998-6008.
[10]	Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative adversarial networks [J]. Communications of the ACM, 2020, 63
(11)	: 139-144.
[11]	Redmon J, Farhadi A. YOLOv3: An incremental improvement [OL]. https://arxiv.org/abs/1804.02767.
[12]	Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real-time object detection [C]. 2016 IEEE Conference
	on Computer Vision and Pattern Recognition (CVPR). June 27-30, 2016, Las Vegas, NV, USA. IEEE, 2016: 779-788.
[13]	Redmon J, Farhadi A. YOLO9000: Better, faster, stronger [C]. 2017 IEEE Conference on Computer Vision and Pattern
	Recognition (CVPR). July 21-26, 2017, Honolulu, HI, USA. IEEE, 2017: 6517-6525.
[14]	Schechner Y Y, Narasimhan S G, Nayar S K. et al. Polarization-based vision through haze [J]. Applied Optics, 2003, 42(3):
51	1-525.
[15]	Zhang H, Patel V M. Densely connected pyramid dehazing network [C]. 2018 IEEE/CVF Conference on Computer Vision
	and Pattern Recognition. June 18-23, 2018, Salt Lake City, UT, USA. IEEE, 2018: 3194-3203.
[16]	Yin X, Liu X M. Multi-task convolutional neural network for pose-invariant face recognition [J]. IEEE Transactions on Image
	Processing: a Publication of the IEEE Signal Processing Society, 2018, 27(2): 964-975.
[17]	Dong H, Pan J S, Xiang L, et al. Multi-scale boosted dehazing network with dense feature fusion [C]. 2020 IEEE/CVF
	Conference on Computer Vision and Pattern Recognition (CVPR). June 13-19, 2020, Seattle, WA, USA. IEEE, 2020: 2154-
	2164.
[18]	Li R D, Pan J S, Li Z C, et al. Single image dehazing via conditional generative adversarial network [C]. 2018 IEEE/CVF
	Conference on Computer Vision and Pattern Recognition. June 18-23, 2018, Salt Lake City, UT, USA. IEEE, 2018: 8202-
	8211.
[19]	Pan J S, Dong J X, Liu Y, et al. Physics-based generative adversarial models for image restoration and beyond [J]. IEEE
	Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(7): 2449-2462.
[20]	Engin D, Genc A, Ekenel H K. Cycle-dehaze: Enhanced CycleGAN for single image dehazing [C]. 2018 IEEE/CVF
	Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). June 18-22, 2018, Salt Lake City, UT,
	USA. IEEE, 2018: 938-9388.
[21]	Wang Z, Bovik A C, Sheikh H R, et al. Image quality assessment: From error visibility to structural similarity [J]. IEEE
	Transactions on Image Processing: A Publication of the IEEE Signal Processing Society, 2004, 13(4): 600-612.