基于深度卷积神经网络的自适应图像去雾算法

doi:10.16180/j.cnki.issn1007-7820.2020.08.012

电子科技 ›› 2020, Vol. 33 ›› Issue (8): 70-73.doi: 10.16180/j.cnki.issn1007-7820.2020.08.012

基于深度卷积神经网络的自适应图像去雾算法

何宜鸿,李彦锋,黄树恺,谭万钏

广东工业大学计算机学院,广东广州 510006

收稿日期:2019-07-11 出版日期:2020-08-15 发布日期:2020-08-24
作者简介:何宜鸿(1999-),男,本科。研究方向:计算机视觉。
基金资助:
广东工业大学大学生创新训练项目(XJ202011845349)

Adaptive Image Dehazing Algorithm Based on Deep Convolutional Neural Network

HE Yihong,LI Yanfeng,HUANG Shukai,TAN Wanchuan

School of Computer Science,Guangdong University of Technology,Guangzhou 510006,China

Received:2019-07-11 Online:2020-08-15 Published:2020-08-24
Supported by:
Innovation Training Program for College Students of Guangdong University of Technology(XJ202011845349)

摘要/Abstract

摘要：

室外拍摄图像由于受雾气、雾霾、沙尘等大气颗粒杂质的影响呈现出图像灰白化,而现有的图像去雾算法存在过度依赖先验信息、透射率计算不精确的问题。针对上述问题,文中提出了一种基于深度卷积神经网络的自适应图像去雾算法。该算法基于大气散射模型实现了有雾图像的去雾,设计浅层提取、并行提取和深度融合共3个全卷积网络实现图像浅层特征与深层特征的融合,大幅提高了透射率图的准确性。去雾实验测试结果表明,文中所提出的算法对室外露天雾图具有良好的去雾效果,且去雾细节效果更加理想。

关键词: 深度卷积, 神经网络, 图像去雾, 大气散射模型, 室外雾图, 全卷积, 透射率, 细节去雾

Abstract:

Due to the influence of air particles such as fog, haze, dust and so on, the image of outdoor photography was gray and white. However, the existing image de fogging algorithm had the problems of over dependence on prior information and inaccurate transmission calculation. In order to solve the above problems, an adaptive image defogging algorithm based on the depth convolution neural network was proposed in this paper. The algorithm realized the defogging of the foggy image based on the atmospheric scattering model. Three full convolution networks including shallow extraction, parallel extraction and deep fusion, were designed to realize the fusion of the shallow and deep features of the image, which greatly improved the accuracy of the transmittance image. The experimental results showed that the algorithm proposed in this paper had a good defogging effect on outdoor fog map, and the effect of defogging details was better.

Key words: deep convolution, neural network, image defogging, atmospheric scattering model, outdoor fog map, full convolution, transmission, detail dehazing

中图分类号:

TP391.41

何宜鸿,李彦锋,黄树恺,谭万钏. 基于深度卷积神经网络的自适应图像去雾算法[J]. 电子科技, 2020, 33(8): 70-73.

HE Yihong,LI Yanfeng,HUANG Shukai,TAN Wanchuan. Adaptive Image Dehazing Algorithm Based on Deep Convolutional Neural Network[J]. Electronic Science and Technology, 2020, 33(8): 70-73.

图/表 5

参考文献 11

[1]	倪继平. 基于机器学习的无人驾驶技术研究[J]. 计算机与网络, 2018,44(6):68-70. doi: 10.1109/MC.2011.40
	Ni Jiping. Research of driverless technology based on machine learning[J]. Computer & Network, 2018,44(6):68-70. doi: 10.1109/MC.2011.40
[2]	杨涵. 基于深度学习的人脸识别算法研究[J]. 电子设计工程, 2019,27(19):155-159.
	Yang Han. Research on face recognition algorithm based on deep learning[J]. Electronic Design Engineering, 2019,27(19):155-159.
[3]	王锋, 李培灵, 梁义涛, 等. 基于暗通道先验的图像去雾改进算法[J]. 科学技术与工程, 2019,19(33):308-313.
	Wang Feng, Li Peiling, Liang Yitao, et al. Improved image dehazing algorithm based on dark channel prior[J]. Science Technology and Engineering, 2019,19(33):308-313.
[4]	吴迪, 朱青松. 图像去雾的最新研究进展[J]. 自动化学报, 2015,41(2):221-239. doi: 10.16383/j.aas.2015.c131137
	Wu Di, Zhu Qingsong. Recent research progress of image dehazing[J]. Journal of Automation, 2015,41(2):221-239. doi: 10.16383/j.aas.2015.c131137
[5]	He K M, Sun J, Tang X O. Guided image filtering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013,25(6):1397-1409.
[6]	吴国鹏, 李盾, 何宇龙, 等. 一种基于暗通道理论的海上图像去雾算法[J]. 光学与光电技术, 2019,17(6):45-50.
	Wu Guopeng, Li Dun, He Yulong, et al. Adehazing algorithm of marine images based on improved dark channel prior theory[J]. Optics & Optoelectronic Technology, 2019,17(6):45-50.
[7]	Meng G F, Wang Y, Duan J Y, et al. Efficient image dehazing with boundary constraint and contextual regularization[C]. Sydney:IEEE International Conference on Computer Vision, IEEE, 2013.
[8]	姬旭丽, 王亚刚, 车芳. 基于深度学习的图像去雾霾研究[J]. 信息技术, 2018,42(8):106-110.
	Ji Xuli, Wang Yagang, Che Fang. Images study on the dehaze based on deep learning[J]. Information Technology, 2018,42(8):106-110.
[9]	乔伟伟. 基于随机森林和核回归的图像去雾研究[D]. 苏州:苏州大学, 2017.
	Qiao Weiwei. Research on image defogging based on random forest and kemelrgression[D]. Suzhou:Soochow University, 2017.
[10]	Ren W, Liu S, Zhang H, et al. Single image dehazing via multi-scale convolutional neural networks[C]. Berlin: Computer Vision-ECCV,Springer International Publishing, 2016.
[11]	Cox L J. Optics of the atmosphere-scattering by molecules and particles[J]. Optica Acta:International Journal of Optics, 1997,27(7):779-798.

[1]	李辉,王一丞. 基于CNNCIFG-Attention模型的文本情感分类[J]. 电子科技, 2022, 35(2): 46-51.
[2]	程晓雅,张雷. 基于深度学习的有遮挡人脸识别方法研究[J]. 电子科技, 2022, 35(1): 35-39.
[3]	战荫伟,朱百万,杨卓. 一种车脸识别算法的研究与应用[J]. 电子科技, 2021, 34(8): 1-7.
[4]	廖璐明,张伟. 基于改进VGG16网络的混合批量训练交通标志识别[J]. 电子科技, 2021, 34(8): 8-13.
[5]	司明明,陈玮,胡春燕,尹钟. 融合Resnet50和U-Net的眼底彩色血管图像分割[J]. 电子科技, 2021, 34(8): 19-24.
[6]	马立新,豆晨飞,宋晨灿,杨天笑. 基于特征融合卷积神经网络的绝缘子无损检测[J]. 电子科技, 2021, 34(7): 26-30.
[7]	顾伟,李菲菲,陈虬. 基于多特征融合的行人检测方法[J]. 电子科技, 2021, 34(5): 29-34.
[8]	孟晓姣,张世巍,李小健,李敏玥,宋丙鑫,路宏敏. 一种基于BP神经网络的车载通信设备性能评估方法[J]. 电子科技, 2021, 34(5): 24-28.
[9]	张仕森,孙宪坤,尹玲,李世玺. 基于神经网络的文本标题生成原型系统设计[J]. 电子科技, 2021, 34(5): 35-41.
[10]	薛永杰,巨志勇. 基于改进AlexNet的鱼类识别算法[J]. 电子科技, 2021, 34(4): 12-17.
[11]	闫书豪,乔美英. 基于一维WConv-BiLSTM的轴承故障诊断算法[J]. 电子科技, 2021, 34(4): 75-82.
[12]	窦征立,王亚刚. 基于Elman神经网络的联合循环机组燃烧室温度模型建模[J]. 电子科技, 2021, 34(3): 60-64.
[13]	葛靖,刘子龙. 基于CNN和LSTM的睡眠呼吸暂停检测算法[J]. 电子科技, 2021, 34(2): 21-26.
[14]	朱斌,刘子龙. 基于新型初始模块的卷积神经网络图像分类方法[J]. 电子科技, 2021, 34(2): 52-56.
[15]	周小波,邹任玲,卢旭华,王海滨,张俊翔. 基于人工神经网络的表面肌电信号分类器研究进展[J]. 电子科技, 2021, 34(2): 62-67.

基于深度卷积神经网络的自适应图像去雾算法

Adaptive Image Dehazing Algorithm Based on Deep Convolutional Neural Network

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 5

参考文献 11

相关文章 15

Metrics

本文评价

推荐阅读 10