PDF下载分享

[1]陈留,杨笔锋,谢欢,等.基于纹理和SVM的地面凝结现象观测方法研究[J].成都信息工程大学学报,2022,37(06):622-626.[doi:10.16836/j.cnki.jcuit.2022.06.002]
　CHEN Liu,YANG Bifeng,XIE Huan,et al.A Texture-based SVM Observation of Ground Condensation Phenomenon[J].Journal of Chengdu University of Information Technology,2022,37(06):622-626.[doi:10.16836/j.cnki.jcuit.2022.06.002]

点击复制

基于纹理和SVM的地面凝结现象观测方法研究

成都信息工程大学学报[ISSN:1006-6977/CN:61-1281/TN] 卷: 37 期数: 2022年06期页码: 622-626 栏目: 电子信息科学与技术出版日期: 2022-12-30

Title:: A Texture-based SVM Observation of Ground Condensation Phenomenon

文章编号:: 2096-1618(2022)06-0622-05

作者:: 陈留¹; 杨笔锋¹; 2; 谢欢¹; 马尚昌¹; 2; (1.成都信息工程大学电子工程学院,四川成都 610225; 2.中国气象局大气探测重点开放实验室,四川成都 610225)

Author(s):: CHEN Liu¹; YANG Bifeng¹; 2; XIE Huan¹; MA Shangchang¹; 2; (1.College of Electronical Engineering,Chengdu University of Information Technology, Chengdu 610225,China; 2.The Key Laboratory of China Meteorological Administration, Chengdu 610225,China)

关键词:: 边缘检测; 纹理特征; 支持向量机; 图像识别

Keywords:: edge detection; texture feature; support vector machines; image identification

分类号:: TN911.73

DOI:: 10.16836/j.cnki.jcuit.2022.06.002

文献标志码:: A

摘要:: 地面凝结现象传统观测方法为人工观测,存在观测时效性差和主观性强等问题,影响观测数据质量。通过图像处理技术识别天气现象是当前研究的热点,但是霜露容易受到温度和光照的影响而消失,霜露在图像中的分布特征容易受到干扰,提取特征难度较大,因此该方法在地面凝结现象识别率方面还有较大的提升空间。对此提出一种高清CCD拍照自动识别地面凝结现象的方案,首先提取图片的感兴趣区域,然后将感兴趣区域通过Canny边缘检测提取纹理特征等一系列处理,最终利用支持向量机将结霜、结露、干燥3种现象进行识别。这种方法不受光线影响,可以同时支持在白天和夜晚观测。通过和人工观测对比,最终综合识别准确率为86.5%。

Abstract:: The traditional observation method of ground condensation phenomenon is manual observation, which has problems such as poor observation timeliness and strong subjectivity, which affect the quality of observation data.Recognizing weather phenomena through image processing technology is a current popular trend, butdew and frost are easy to disappear due to temperature and light. The distribution of frost and dew in the image is disturbed easily.So, it is difficult to extract features. Therefore, this method still has more room for improvement in the recognition rate of ground condensation phenomena. This article introduces a high-definition CCD camerato automatically identify the ground condensation phenomenon.First, extract the interested region of the picture, then, extract the feature of the interested region by canny edge detection. Finally, identify the three phenomena of frost, dew, and dry by the Support Vector Machine.This method is not only unaffected by light, but also supports observation during the day and night.Compared with manual observation, the final comprehensive recognition accuracy rate is 86.5%.

参考文献/References:

[1] 温显罡,马舒庆,杜波,等.草温与露点温度对结露和结霜指示性分析[J].气象,2012,38(6):745-750.
[2] 朱磊,曹治国,肖阳,等.基于流型学习的地面结霜现象检测[J].计算机应用,2015,35(3):854-857.
[3] 蒲晓虎,马舒庆,梁丽,等.霜、露自动化观测的研究[J].气象,2015,41(9):1152-1158.
[4] 李肖霞,马舒庆,吴可军,等.结露自动化观测装置及试验研究[J].气象,2012,38(4):501-507.
[5] Jerzy Weremezuk.Dew/Frost Point Recognition with Fingerprint Sensor[J]. Instrumentation and Measurement Technology Conference-IMTC 2007 Warsaw Poland,2007.
[6] 周忠义.基于深度学习的地表凝结图像分类研究[D].武汉:武汉科技大学,2018.
[7] 秦雪,王欢,严晔,等.基于图像边缘检测的图像识别算法[J].长春理工大学学报(自然科学版),2015,38(4):127-131.
[8] 姚洋,彭道刚,王志萍.基于改进canny检测与Hough变换的仪表图像识别算法[J].上海电力大学学报,2020,36(2):183-189.
[9] 王植,贺赛先.一种基于canny理论的自适应边缘检测方法[J].中国图像图形学报,2004,9(8):957-962.
[10] 杨红云,黄琼,孙爱珍,等.基于卷积神经网络和支持向量机的水稻种子图像分类识别[J].中国粮油学报,2021,36(12):144-150.
[11] 彭荣杰,彭亚雄,陆安江.基于改进PCA+SVM的人脸识别系统[J].电子科技,2021,34(12):56-61.
[12] 陈荣,李旺,周文玉.基于形状特征和支持向量机(SVM)的茶叶病害识别方法[J].贵州农业科学,2021,49(4):53-59.
[13] 段利超.基于纹理特征融合和SVM的人脸识别算法研究[D].重庆:重庆大学,2015.

备注/Memo

收稿日期:2021-12-17

更新日期/Last Update: 2022-12-30