图像能量与对比度的霾检测算法

孔明; 杨天琪; 单良; 郭天太; 王道档; 徐良

doi:10.3788/IRLA201847.0826001

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

图像能量与对比度的霾检测算法

孔明, 杨天琪, 单良, 郭天太, 王道档, 徐良

文章导航 > 红外与激光工程 > 2018 > 47(8): 826001-0826001(6)

孔明, 杨天琪, 单良, 郭天太, 王道档, 徐良. 图像能量与对比度的霾检测算法[J]. 红外与激光工程, 2018, 47(8): 826001-0826001(6). doi: 10.3788/IRLA201847.0826001

引用本文:

孔明, 杨天琪, 单良, 郭天太, 王道档, 徐良. 图像能量与对比度的霾检测算法[J]. 红外与激光工程, 2018, 47(8): 826001-0826001(6). doi: 10.3788/IRLA201847.0826001

Kong Ming, Yang Tianqi, Shan Liang, Guo Tiantai, Wang Daodang, Xu Liang. Haze detection algorithm based on image energy and contrast[J]. Infrared and Laser Engineering, 2018, 47(8): 826001-0826001(6). doi: 10.3788/IRLA201847.0826001

Citation:

Kong Ming, Yang Tianqi, Shan Liang, Guo Tiantai, Wang Daodang, Xu Liang. Haze detection algorithm based on image energy and contrast[J]. Infrared and Laser Engineering, 2018, 47(8): 826001-0826001(6). doi: 10.3788/IRLA201847.0826001

图像能量与对比度的霾检测算法

doi: 10.3788/IRLA201847.0826001

1.
中国计量大学计量测试工程学院,浙江杭州 310018

基金项目:

国家自然科学基金（51476154，51404223）;浙江省自然科学基金（LY13E060006，LQ14E060003）

详细信息

作者简介:
孔明(1978-),男,教授,博士,主要从事光学测量、精密仪器方面的研究。Email:mkong@cjlu.edu.cn

中图分类号: TP391.4

Haze detection algorithm based on image energy and contrast

1.
School of Measurement and Testing Engineering,China Jiliang University,Hangzhou 310018,China

摘要: 针对目前霾检测方法实时性差且成本较高的问题，提出一种基于图像能量与对比度的霾检测方法。首先，对CMOS相机拍摄的图像进行预处理。由于相机受外界影响会出现轻微摆动，故需对图像进行配准；其次，在图像的关键区域中获取目标与水平天空背景的对比度和图像能量两个特征向量；再次，将对比度、图像能量、环境湿度作为输入，将激光粒子计数器测得的实时PM10浓度作为输出，进行支持向量回归训练，建立图像和PM10浓度间的关系模型；最后，根据得到的模型计算待测图像所对应的PM10浓度。将该方法检测的PM10浓度与激光粒子计数器测得浓度值进行对比，实验表明该方法检测结果的平均相对误差在10%以内，MSE为0.006 2，表明预测值与真值拟合程度较好，模型检测的精度较高。在此基础上增加训练样本可进一步提高模型精度。此外，该方法可针对不同待测环境建立相应的关系模型，具有较强的灵活性。
- 图像处理 /
- 霾 /
- 对比度 /
- PM10 /
- 支持向量回归机
Abstract: In view of the poor real-time performance and high cost of haze detection methods, a method based on contrast and image energy was proposed to detect the haze. Firstly, the images taken by the CMOS camera were preprocessed. The image has some slight swing because the camera has been disturbed by the external environment, so the images were registered. Secondly, in the critical region of the image, two contrast vectors of contrast and image energy were obtained. Thirdly, the contrast, image energy and ambient humidity were taken as input, and the real-time PM10 concentration measured by the laser particle counter was used as the output. The relational model between input and output was constructed by training support vector regression(SVR). Finally, the PM10 concentration of the image was calculated using the model. The PM10 concentration detected by this method was compared with that measured by laser particle counter. The average relative error was less than 10% and MSE was 0.006 2, which indicates that the fitting degree between the predicted value and the true value is good and the accuracy of the model was high. On this basis, increasing the training samples can improve the model accuracy. Moreover, the method can establish the corresponding relation model for different environment to be tested, which has strong flexibility.
- image processing /
- haze /
- contrast /
- PM10 /
- support vector machine for regression

[1]	Wu Dan, Yu Yaxin, Xia Junrong, et al. Long-term variation in haze days and related climatic factors in Nanjing[J].Journal of Nanjing Institute of Meteorology, 2016, 39(2):232-242. (in Chinese)
[2]	Finsy R, Deriemaeker L, Gelad E, et al. Inversion of static light scattering measurements for particle size distributions[J]. Journal of Colloid and Interface Science, 1992, 153(2):337-354.
[3]	Cao Lixia. Research on measurement technology of particle size based on static light scattering[D]. Hangzhou:China Jiliang University, 2015. (in Chinese)
[4]	Fan Jiao, Guo Baofeng, He Hongchang. Retrieval of aerosol optical thickness with MODIS data over Hangzhou[J]. Acta Optica Sinica, 2015, 35(1):9-17. (in Chinese)
[5]	Lu Lipeng, Wang Bin, Liu Hui, et al. Haze pollution level detection method based on image gray differential statistics[J]. Computer Engineering, 2016, 42(1):225-230. (in Chinese)
[6]	Steffens C. Measurement of visibility by photographic photometry[J]. Industrial Engineering Chemistry, 1949, 41(11):2396-2399.
[7]	Han Mingmin. High waility detecting technology based on video images[D]. Beiijng:Beijing Jiaotong University, 2016. (in Chinese)
[8]	Harris C G, Stephens M J. A combined corner and edge detector[C]//Proceedings of the 4th Alvey Vision Conference Manchester, 1988:147-151.
[9]	Xu Xi, Yin Xucheng, Li Yan, et al. Visibility measurement with image understanding[J]. Pattern Recognition and Artificial Intelligence, 2013, 26(6):543-551. (in Chinese)
[10]	Cortes C, Vapnik V. Support-vector networks[J]. Machine Learning, 1995, 20(3):273-297.

[1]	李鹏越, 续欣莹, 唐延东, 张朝霞, 韩晓霞, 岳海峰. 基于并行多轴自注意力的图像去高光算法 . 红外与激光工程, 2024, 53(3): 20230538-1-20230538-11. doi: 10.3788/IRLA20230538
[2]	李居尚, 战荫泽, 张立东. 组合沃拉斯顿棱镜组的目标偏振检测方法 . 红外与激光工程, 2021, 50(11): 20210339-1-20210339-5. doi: 10.3788/IRLA20210339
[3]	薛芳, 陈永权, 段亚轩, 蔺辉, 达争尚. 高能激光测量中的光强探测器面响应校正方法 . 红外与激光工程, 2021, 50(S2): 20210215-1-20210215-8. doi: 10.3788/IRLA20210215
[4]	黄宜帆, 贺岩, 胡善江, 侯春鹤, 朱小磊, 李凯鹏, 刘芳华, 陈勇强, 郭守川. 海洋激光雷达图像处理提取海水深度的方法 . 红外与激光工程, 2021, 50(6): 20211034-1-20211034-8. doi: 10.3788/IRLA20211034
[5]	付伟伟, 黄坤. 基于微纳器件的全光图像处理技术及应用 . 红外与激光工程, 2020, 49(9): 20201040-1-20201040-14. doi: 10.3788/IRLA20201040
[6]	田永胜, 侯金, 龙银福, 唐翠姣, 黄玉春, 杨春勇, 陈少平. 光纤倾斜耦合角度的快速精密图像测量 . 红外与激光工程, 2019, 48(10): 1013001-1013001(8). doi: 10.3788/IRLA201948.1013001
[7]	赵永强, 李宁, 张鹏, 姚嘉昕, 潘泉. 红外偏振感知与智能处理 . 红外与激光工程, 2018, 47(11): 1102001-1102001(7). doi: 10.3788/IRLA201847.1102001
[8]	秦齐, 刘艳, 刘欢欢, 时川, 谭中伟. 图像处理在光纤光斑微位移传感中的应用 . 红外与激光工程, 2018, 47(10): 1022004-1022004(7). doi: 10.3788/IRLA201847.1022004
[9]	骆媛, 张科, 纪明. 增强无人机态势感知的彩色图像融合方法 . 红外与激光工程, 2016, 45(S1): 200-206. doi: 10.3788/IRLA201645.S126002
[10]	张智, 林栩凌, 何红艳. 一种基于量子力学的遥感图像滤波方法研究 . 红外与激光工程, 2016, 45(S2): 140-144. doi: 10.3788/IRLA201645.S226001
[11]	夏彦文, 郑奎兴, 刘华, 赵军普, 彭志涛, 孙志红, 董军, 李森. 纳秒整形激光脉冲波形的高对比度诊断 . 红外与激光工程, 2016, 45(11): 1106002-1106002(6). doi: 10.3788/IRLA201645.1106002
[12]	田昌会, 杨百愚, 蔡明, 范琦, 王斌科, 王伟宇, 屈绍波. 大气背景对红外目标探测的影响 . 红外与激光工程, 2014, 43(2): 438-441.
[13]	马媛花, 胡炳樑, 李然, 孙朗, 孙念, 王峥杰. 采用Gyrator变换的泰伯效应及图像去噪 . 红外与激光工程, 2014, 43(2): 665-670.
[14]	田昌会, 蔡明, 杨百愚, 范琦, 王斌科, 王伟宇, 屈绍波. 星载红外探测对比度的计算与分析 . 红外与激光工程, 2014, 43(3): 737-741.
[15]	任广辉, 王刚毅, 金炎胜. 利用FPGA的高性能向导滤波器设计 . 红外与激光工程, 2013, 42(2): 537-542.
[16]	赵大鹏, 时家明, 汪家春, 王启超, 王伟. 偏振热成像系统的作用距离分析 . 红外与激光工程, 2013, 42(5): 1146-1152.
[17]	韩艳丽, 王铎, 张健, 樊利恒, 孙腾飞. 近红外多视场白天测星分析 . 红外与激光工程, 2013, 42(8): 2202-2208.
[18]	张法全, 王国富, 曾庆宁, 叶金才. 利用重心原理的图像目标最小外接矩形快速算法 . 红外与激光工程, 2013, 42(5): 1382-1387.
[19]	李运达, 李琦, 刘正君, 王骐. 太赫兹计算机辅助层析图像重构算法仿真研究 . 红外与激光工程, 2013, 42(5): 1228-1235.
[20]	董建婷, 杨小乐, 董杰. 面阵CMOS图像传感器性能测试及图像处理 . 红外与激光工程, 2013, 42(12): 3396-3401.

点击查看大图

计量

文章访问数: 515
HTML全文浏览量: 82
PDF下载量: 35
被引次数: 0

图像能量与对比度的霾检测算法

doi: 10.3788/IRLA201847.0826001

1. 中国计量大学计量测试工程学院,浙江杭州 310018

作者简介:
孔明(1978-),男,教授,博士,主要从事光学测量、精密仪器方面的研究。Email:mkong@cjlu.edu.cn

收稿日期: 2018-03-13
修回日期: 2018-04-17
刊出日期: 2018-08-25

基金项目:

国家自然科学基金（51476154，51404223）;浙江省自然科学基金（LY13E060006，LQ14E060003）

中图分类号: TP391.4

关键词:

摘要: 针对目前霾检测方法实时性差且成本较高的问题，提出一种基于图像能量与对比度的霾检测方法。首先，对CMOS相机拍摄的图像进行预处理。由于相机受外界影响会出现轻微摆动，故需对图像进行配准；其次，在图像的关键区域中获取目标与水平天空背景的对比度和图像能量两个特征向量；再次，将对比度、图像能量、环境湿度作为输入，将激光粒子计数器测得的实时PM10浓度作为输出，进行支持向量回归训练，建立图像和PM10浓度间的关系模型；最后，根据得到的模型计算待测图像所对应的PM10浓度。将该方法检测的PM10浓度与激光粒子计数器测得浓度值进行对比，实验表明该方法检测结果的平均相对误差在10%以内，MSE为0.006 2，表明预测值与真值拟合程度较好，模型检测的精度较高。在此基础上增加训练样本可进一步提高模型精度。此外，该方法可针对不同待测环境建立相应的关系模型，具有较强的灵活性。

English Abstract

Haze detection algorithm based on image energy and contrast

1. School of Measurement and Testing Engineering,China Jiliang University,Hangzhou 310018,China

Received Date: 2018-03-13
Rev Recd Date: 2018-04-17
Publish Date: 2018-08-25

Keywords:

Abstract: In view of the poor real-time performance and high cost of haze detection methods, a method based on contrast and image energy was proposed to detect the haze. Firstly, the images taken by the CMOS camera were preprocessed. The image has some slight swing because the camera has been disturbed by the external environment, so the images were registered. Secondly, in the critical region of the image, two contrast vectors of contrast and image energy were obtained. Thirdly, the contrast, image energy and ambient humidity were taken as input, and the real-time PM10 concentration measured by the laser particle counter was used as the output. The relational model between input and output was constructed by training support vector regression(SVR). Finally, the PM10 concentration of the image was calculated using the model. The PM10 concentration detected by this method was compared with that measured by laser particle counter. The average relative error was less than 10% and MSE was 0.006 2, which indicates that the fitting degree between the predicted value and the true value is good and the accuracy of the model was high. On this basis, increasing the training samples can improve the model accuracy. Moreover, the method can establish the corresponding relation model for different environment to be tested, which has strong flexibility.