基于改进Faster-RCNN模型的粘虫板图像昆虫识别与计数
|
摘要
针对传统机器学习采用人工提取特征方法时,由于人为主观性而影响昆虫识别效果与计数准确性的问题,采用图像特征自动提取方法,将深度学习目标检测模型引入昆虫的识别与计数领域,对Faster-RCNN目标检测模型进行改进:针对昆虫体积小,图像分辨率较低的特点,用网络深度更深,运算量更小的深度残差网络(ResNet50)代替原来的VGG16,以提取更加丰富的特征;针对部分昆虫密集的特点,用Soft-NMS算法代替传统的非极大值抑制(NMS)算法,以减少密集区域的漏检。结果表明:改进后Faster-RCNN模型的检测准确率达到90.7%,较未改进的Faster-RCNN模型提高了4.2%,可以运用于昆虫的分类计数。利用深度学习目标检测模型进行昆虫识别与计数较传统的昆虫识别与计数方法更加方便,能够将昆虫的识别、定位和计数融为一体。
In view of the traditional machine learning using artificial extraction feature method,due to the subjectivity of humans affecting the effect of insect recognition and the accuracy of counting,this study uses the automatic extraction of image features and introduces the deep learning target detection model into insect recognition.Compared with the counting field,the Faster-RCNN target detection model is improved:In view of the small insect size and low image resolution,the deep residual network(ResNet50) with deeper network depth and smaller computational capacity is used rather than the original VGG16.To extract more abundant features.Because of the high density of some insects,the Soft-NMS algorithm is used to replace the traditional non-maximum value suppression(NMS) algorithm in order to reduce the missed detection in dense areas.The results show that the improved Faster-RCNN model has an accuracy of 90.7%,which is 4.2% higher than that of the unmodified Faster-RCNN model,and can be applied to the classification and counting of insects.The use of deep learning target detection model for insect identification and counting is more convenient than traditional insect identification and counting methods.It can integrate insect identification,localization and counting.
In view of the traditional machine learning using artificial extraction feature method,due to the subjectivity of humans affecting the effect of insect recognition and the accuracy of counting,this study uses the automatic extraction of image features and introduces the deep learning target detection model into insect recognition.Compared with the counting field,the Faster-RCNN target detection model is improved:In view of the small insect size and low image resolution,the deep residual network(ResNet50) with deeper network depth and smaller computational capacity is used rather than the original VGG16.To extract more abundant features.Because of the high density of some insects,the Soft-NMS algorithm is used to replace the traditional non-maximum value suppression(NMS) algorithm in order to reduce the missed detection in dense areas.The results show that the improved Faster-RCNN model has an accuracy of 90.7%,which is 4.2% higher than that of the unmodified Faster-RCNN model,and can be applied to the classification and counting of insects.The use of deep learning target detection model for insect identification and counting is more convenient than traditional insect identification and counting methods.It can integrate insect identification,localization and counting.
引文
[1] 黄世国.基于图像的昆虫识别关键技术研究[D].西安:西北大学,2008Huang S G.Research on the key technologies of image-based insectrecognition[D].Xi’an:Northwest University,2008 (in Chinese)
[2] 姚青,吕军,杨保军,薛杰,郑宏海,唐健.基于图像的昆虫自动识别与计数研究进展[J].中国农业科学,2011,44(14):2886-2899Yao Q,Lü J,Yang B J,Xue J,Zheng H H,Tang J.Progress in research on digital image processing technology for automatic insect identification and counting[J].Scientia Agricultura Sinica,2011,44(14):2886-2899 (in Chinese)
[3] Zhu L Q,Zhang Z.Auto-classification of insect images based on color histogram and GLCM[C].In:2010 Seventh International Conference on Fuzzy Systems and Knowledge Discovery.Yantai:IEEE,2010:2589-2593
[4] 范艳峰,甄彤.谷物害虫检测与分类识别技术的研究及应用[J].计算机工程,2005,31(12):187-189Fan Y F,Zhen T.Research and application of grain pests detection and classificationtechnology[J].Computer Engineering,2005,31(12):187-189 (in Chinese)
[5] 刘芳,沈佐锐,张建伟,杨红珍.基于颜色特征的昆虫自动鉴定方法[J].昆虫知识,2008,45(1):150-153Liu F,Shen Z R,Zhang J W,Yang H Z.Automatic insect identification based on color characters[J].Chinese Bulletin of Entomology,2008,45(1):150-153 (in Chinese)
[6] Cho J,Choi J,Qiao M,Ji C W,Kim H Y,Uhm K B,Chon T S.Automatic identification of whiteflies,Aphids and thrips in greenhouse based on image analysis[J].International Journal of Mathematics & Computers in Simulation,2007,1(1):46-53
[7] Zhao J,Cheng X P.Field pest identification by an improvedgabor texture segmentation scheme[J].New Zealand Journal of Agricultural Research,2007,50(5):719-723
[8] 杨国国,鲍一丹,刘子毅.基于图像显著性分析与卷积神经网络的茶园害虫定位与识别[J].农业工程学报,2017,33(6):156-162Yang G G,Bao Y D,Liu Z Y.Localization and recognition of pests in Tea plantation based on image saliency analysis and convolutional neural network[J].Transactions of the Chinese Society of Agricultural Engineering,2017,33(6):156-162 (in Chinese)
[9] 王利强.点特征配准算法及其在储粮害虫种类识别中的应用研究[D].郑州:河南工业大学,2011Wang L Q.Study point feature matching algorithm and its application in stored grain pest speciesidentification[D].Zhengzhou:Henan University of Technology,2011 (in Chinese)
[10] Hinton G E.Reducing the dimensionality of data with neuralnetworks[J].Science,2006,313(5786):504-507
[11] 张文达,许悦雷,倪嘉成,马时平,史鹤欢.基于多尺度分块卷积神经网络的图像目标识别算法[J].计算机应用,2016,36(4):1033-1038Zhang W D,Xu Y L,Ni J C,Ma S P,Shi H H.Image target recognition method based on multi-scale block convolutional neural network[J].Journal of Computer Applications,2016,36(4):1033-1038 (in Chinese)
[12] 田壮壮,占荣辉,胡杰民,张军.基于卷积神经网络的SAR图像目标识别研究[J].雷达学报,2016,5(3):320-325Tian Z Z,Zhan R H,Hu J M,Zhang J.SAR ATR based on convolutional neural network[J].Journal of Radars,2016,5(3):320-325 (in Chinese)
[13] 黄洁,姜志国,张浩鹏,姚远.基于卷积神经网络的遥感图像舰船目标检测[J].北京航空航天大学学报,2017,43(9):1841-1848Huang J,Jiang Z G,Zhang H P,Yao Y.Ship object detection in remote sensing images using convolutional neuralnetworks[J].Journal of Beijing University of Aeronautics and Astronautics,2017,43(9):1841-1848 (in Chinese)
[14] 周俊宇,赵艳明.卷积神经网络在图像分类和目标检测应用综述[J].计算机工程与应用,2017,53(13):34-41Zhou J Y,Zhao Y M.Application of convolution neural network in image classification and object detection[J].Computer Engineering and Applications,2017,53(13):34-41 (in Chinese)
[15] Ren S Q,He K M,Girshick R,Sun J.Faster R-CNN:Towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(6):1137-1149
[16] Redmon J,Divvala S,Girshick R,Farhadi A.You only look once:Unified,real-time object detection[C].In:2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:779-788
[17] Liu W,Anguelov D,Erhan D,Szegedy C,Reed S,Fu C Y,Berg A C.SSD:single shot multibox detector[C].In:Computer Vision-ECCV 2016.Cham:Springer International Publishing,2016:21-37
[18] 车广富.基于深度学习的复杂场景行人计数方法的研究[D].北京:北京邮电大学,2017Che G F.Research on pedestrian counting method based on deep learning in complex scenes[D].Beijing:Beijing University of Posts and Telecommunications,2017 (in Chinese)
[19] 贾洪飞.基于深度学习的白细胞分类计数的研究[D].深圳:深圳大学,2017Jia H F.The Research of leukocytes classification based on deep learning[D].Shenzhen:Shenzhen University,2017 (in Chinese)
[20] 罗金燕,周艳孔,李惠明.蔬菜大棚植物诱源黄板诱虫效果研究[J].中国植保导刊,2010,30(11):23-25Luo J,Zhou Y K,Li H M.Study on insect trapping efficiency of yellow board with plant attractant in vegetable greenhouse[J].China Plant Protection,2010,30(11):23-25 (in Chinese)
[21] 董宁禹,刘占卿,赵世民,刘欢欢,王枫,刘涛,马玉欣,蒋士君.太阳能杀虫灯和诱虫黄板绿色防控技术在烟草生产上的应用效果[J].河南农业科学,2015,44(8):83-86Dong N Y,Liu Z Q,Zhao S M,Liu H H,Wang F,Liu T,Ma Y X,Jiang S J.Application effects of green prevention and control technology of solar insecticidal lamp and yellow template in tobacco production[J].Journal of Henan Agricultural Sciences,2015,44(8):83-86 (in Chinese)
[22] He K M,Zhang X Y,Ren S Q,Sun J.Deep residual learning for image recognition[C].In:2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Computer Society,2016:770-778
[23] Simonyan K,Zisserman A.Very deep convolutional networks for large-scale image recognition[J].Computer Science,2014:1-14
[24] Bodla N,Singh B,Chellappa R,Davis L S.Soft-NMS improving object detection with one line of code[C].In:2017 IEEE International Conference on Computer Vision (ICCV).Venice:IEEE,2017:5562-5570
[25] 温明莉,赵轩,蔡梦倩.基于深度学习的端到端验证码识别[J].无线互联科技,2017(14):85-86Wen M L,Zhao X,Cai M Q.End-to-end verification code identification based on depth learning[J].Wireless Internet Technology,2017(14):85-86 (in Chinese)
[2] 姚青,吕军,杨保军,薛杰,郑宏海,唐健.基于图像的昆虫自动识别与计数研究进展[J].中国农业科学,2011,44(14):2886-2899Yao Q,Lü J,Yang B J,Xue J,Zheng H H,Tang J.Progress in research on digital image processing technology for automatic insect identification and counting[J].Scientia Agricultura Sinica,2011,44(14):2886-2899 (in Chinese)
[3] Zhu L Q,Zhang Z.Auto-classification of insect images based on color histogram and GLCM[C].In:2010 Seventh International Conference on Fuzzy Systems and Knowledge Discovery.Yantai:IEEE,2010:2589-2593
[4] 范艳峰,甄彤.谷物害虫检测与分类识别技术的研究及应用[J].计算机工程,2005,31(12):187-189Fan Y F,Zhen T.Research and application of grain pests detection and classificationtechnology[J].Computer Engineering,2005,31(12):187-189 (in Chinese)
[5] 刘芳,沈佐锐,张建伟,杨红珍.基于颜色特征的昆虫自动鉴定方法[J].昆虫知识,2008,45(1):150-153Liu F,Shen Z R,Zhang J W,Yang H Z.Automatic insect identification based on color characters[J].Chinese Bulletin of Entomology,2008,45(1):150-153 (in Chinese)
[6] Cho J,Choi J,Qiao M,Ji C W,Kim H Y,Uhm K B,Chon T S.Automatic identification of whiteflies,Aphids and thrips in greenhouse based on image analysis[J].International Journal of Mathematics & Computers in Simulation,2007,1(1):46-53
[7] Zhao J,Cheng X P.Field pest identification by an improvedgabor texture segmentation scheme[J].New Zealand Journal of Agricultural Research,2007,50(5):719-723
[8] 杨国国,鲍一丹,刘子毅.基于图像显著性分析与卷积神经网络的茶园害虫定位与识别[J].农业工程学报,2017,33(6):156-162Yang G G,Bao Y D,Liu Z Y.Localization and recognition of pests in Tea plantation based on image saliency analysis and convolutional neural network[J].Transactions of the Chinese Society of Agricultural Engineering,2017,33(6):156-162 (in Chinese)
[9] 王利强.点特征配准算法及其在储粮害虫种类识别中的应用研究[D].郑州:河南工业大学,2011Wang L Q.Study point feature matching algorithm and its application in stored grain pest speciesidentification[D].Zhengzhou:Henan University of Technology,2011 (in Chinese)
[10] Hinton G E.Reducing the dimensionality of data with neuralnetworks[J].Science,2006,313(5786):504-507
[11] 张文达,许悦雷,倪嘉成,马时平,史鹤欢.基于多尺度分块卷积神经网络的图像目标识别算法[J].计算机应用,2016,36(4):1033-1038Zhang W D,Xu Y L,Ni J C,Ma S P,Shi H H.Image target recognition method based on multi-scale block convolutional neural network[J].Journal of Computer Applications,2016,36(4):1033-1038 (in Chinese)
[12] 田壮壮,占荣辉,胡杰民,张军.基于卷积神经网络的SAR图像目标识别研究[J].雷达学报,2016,5(3):320-325Tian Z Z,Zhan R H,Hu J M,Zhang J.SAR ATR based on convolutional neural network[J].Journal of Radars,2016,5(3):320-325 (in Chinese)
[13] 黄洁,姜志国,张浩鹏,姚远.基于卷积神经网络的遥感图像舰船目标检测[J].北京航空航天大学学报,2017,43(9):1841-1848Huang J,Jiang Z G,Zhang H P,Yao Y.Ship object detection in remote sensing images using convolutional neuralnetworks[J].Journal of Beijing University of Aeronautics and Astronautics,2017,43(9):1841-1848 (in Chinese)
[14] 周俊宇,赵艳明.卷积神经网络在图像分类和目标检测应用综述[J].计算机工程与应用,2017,53(13):34-41Zhou J Y,Zhao Y M.Application of convolution neural network in image classification and object detection[J].Computer Engineering and Applications,2017,53(13):34-41 (in Chinese)
[15] Ren S Q,He K M,Girshick R,Sun J.Faster R-CNN:Towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(6):1137-1149
[16] Redmon J,Divvala S,Girshick R,Farhadi A.You only look once:Unified,real-time object detection[C].In:2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:779-788
[17] Liu W,Anguelov D,Erhan D,Szegedy C,Reed S,Fu C Y,Berg A C.SSD:single shot multibox detector[C].In:Computer Vision-ECCV 2016.Cham:Springer International Publishing,2016:21-37
[18] 车广富.基于深度学习的复杂场景行人计数方法的研究[D].北京:北京邮电大学,2017Che G F.Research on pedestrian counting method based on deep learning in complex scenes[D].Beijing:Beijing University of Posts and Telecommunications,2017 (in Chinese)
[19] 贾洪飞.基于深度学习的白细胞分类计数的研究[D].深圳:深圳大学,2017Jia H F.The Research of leukocytes classification based on deep learning[D].Shenzhen:Shenzhen University,2017 (in Chinese)
[20] 罗金燕,周艳孔,李惠明.蔬菜大棚植物诱源黄板诱虫效果研究[J].中国植保导刊,2010,30(11):23-25Luo J,Zhou Y K,Li H M.Study on insect trapping efficiency of yellow board with plant attractant in vegetable greenhouse[J].China Plant Protection,2010,30(11):23-25 (in Chinese)
[21] 董宁禹,刘占卿,赵世民,刘欢欢,王枫,刘涛,马玉欣,蒋士君.太阳能杀虫灯和诱虫黄板绿色防控技术在烟草生产上的应用效果[J].河南农业科学,2015,44(8):83-86Dong N Y,Liu Z Q,Zhao S M,Liu H H,Wang F,Liu T,Ma Y X,Jiang S J.Application effects of green prevention and control technology of solar insecticidal lamp and yellow template in tobacco production[J].Journal of Henan Agricultural Sciences,2015,44(8):83-86 (in Chinese)
[22] He K M,Zhang X Y,Ren S Q,Sun J.Deep residual learning for image recognition[C].In:2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Computer Society,2016:770-778
[23] Simonyan K,Zisserman A.Very deep convolutional networks for large-scale image recognition[J].Computer Science,2014:1-14
[24] Bodla N,Singh B,Chellappa R,Davis L S.Soft-NMS improving object detection with one line of code[C].In:2017 IEEE International Conference on Computer Vision (ICCV).Venice:IEEE,2017:5562-5570
[25] 温明莉,赵轩,蔡梦倩.基于深度学习的端到端验证码识别[J].无线互联科技,2017(14):85-86Wen M L,Zhao X,Cai M Q.End-to-end verification code identification based on depth learning[J].Wireless Internet Technology,2017(14):85-86 (in Chinese)