基于全卷积神经网络的林区航拍图像虫害区域识别方法
|
摘要
针对航拍林区虫害图像的虫害区域不规则和传统识别方法泛化能力差的问题,提出一种基于全卷积神经网络(Fully convolution networks,FCN)的虫害区域识别方法。采用八旋翼无人机航拍虫害林区、获取林区虫害图像,并对虫害区域进行像素级标注,用于模型训练;将VGG16模型的全连接层替换为卷积层,并通过上采样实现端到端的学习;使用预训练的卷积层参数,提升模型收敛速度;采用跳跃结构融合多种特征信息,有效提升识别精度,并通过该方法构造了5种全卷积神经网络。试验表明,针对林区航拍虫害图像,FCN-2s在5种全卷积神经网络中区域识别精度最高,其像素准确率为97. 86%,平均交并比为79. 49%,单幅分割时间为4. 31 s。该方法与K-means、脉冲耦合神经网络、复合梯度分水岭算法相比,像素准确率分别高出44. 93、20. 73、6. 04个百分点,平均交并比分别高出50. 19、35. 67、18. 86个百分点,单幅分割时间分别缩短47. 54、19. 70、11. 39 s,可以实现林区航拍图像的虫害区域快速准确识别,为林业虫害监测和防治提供参考。
Aiming at the problem of irregularity of the pest area in the aerial images taken over forest area( discussed as forestry pest images in the following) and the poor generalization ability of traditional recognition method,a method for pest image segmentation based on full convolution network was proposed to realize automatic recognition of pest area. Firstly,the insect image of the forest area was needed to be obtained by using the eight-rotor UAV aerial photograph technique over the pest forest area,and the pest area was marked with pixels for model training. Secondly,the full connection layer of the VGG16 model was replaced with the convolutional layer,and an end-to-end study was used by implementing up sampling; and then the pre-training convolutional layer parameters were employed to improve the convergence speed of the model; finally,the skip layer was used to fuse a variety of feature information,which effectively improved the recognition accuracy,and five convolutional networks was constructed by this method. Experiment results showed that FCN-2 s had the highest recognition accuracy among the five full-convolution networks for forestry pest images. The pixel accuracy of the segmentation results was97. 86%,the mean crossover ratio was 79. 49%,and the segmentation time for single image was 4. 31 s.Compared with K-means,pulse coupled neural network and composite gradient watershed algorithm,its pixel accuracy was higher by 44. 93,20. 73 and 6. 04 percentage points, respectively, the mean intersection over union towered above 50. 19,35. 67 and 18. 86 percentage points,and its segmentation time for single image was reduced by 47. 54 s,19. 70 s and 11. 39 s,respectively. This method can realize the rapid and accurate recognition of pest area in aerial forest areas,which provide a basis for pest detection and prevention in forest areas.
Aiming at the problem of irregularity of the pest area in the aerial images taken over forest area( discussed as forestry pest images in the following) and the poor generalization ability of traditional recognition method,a method for pest image segmentation based on full convolution network was proposed to realize automatic recognition of pest area. Firstly,the insect image of the forest area was needed to be obtained by using the eight-rotor UAV aerial photograph technique over the pest forest area,and the pest area was marked with pixels for model training. Secondly,the full connection layer of the VGG16 model was replaced with the convolutional layer,and an end-to-end study was used by implementing up sampling; and then the pre-training convolutional layer parameters were employed to improve the convergence speed of the model; finally,the skip layer was used to fuse a variety of feature information,which effectively improved the recognition accuracy,and five convolutional networks was constructed by this method. Experiment results showed that FCN-2 s had the highest recognition accuracy among the five full-convolution networks for forestry pest images. The pixel accuracy of the segmentation results was97. 86%,the mean crossover ratio was 79. 49%,and the segmentation time for single image was 4. 31 s.Compared with K-means,pulse coupled neural network and composite gradient watershed algorithm,its pixel accuracy was higher by 44. 93,20. 73 and 6. 04 percentage points, respectively, the mean intersection over union towered above 50. 19,35. 67 and 18. 86 percentage points,and its segmentation time for single image was reduced by 47. 54 s,19. 70 s and 11. 39 s,respectively. This method can realize the rapid and accurate recognition of pest area in aerial forest areas,which provide a basis for pest detection and prevention in forest areas.
引文
[1]许章华,李聪慧,刘健,等.马尾松毛虫害等级的Fisher判别分析[J/OL].农业机械学报,2014,45(6):275-283.XU Zhanghua,LI Conghui,LIU Jian,et al. Fisher discriminant analysis of dendrolimus punctatus walker pest levels[J/OL].Transactions of the Chinese Society for Agricultural Machinery,2014,45(6):275-283. http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20140642&flag=1&journal_id=jcsam. DOI:10. 6041/j. issn. 1000-1298. 2014. 06. 042.(in Chinese)
[2]宋玉双,苏宏钧,于海英,等. 2006—2010年我国林业有害生物灾害损失评估[J].中国森林病虫,2011(6):1-4.SONG Yushuang,SU Hongjun,YU Haiying,et al. Evaluation of economic losses caused by forest pest disasters between 2006and 2010 in China[J]. Forest Pest&Disease,2011(6):1-4.(in Chinese)
[3]李根秋.加快林业信息化建设促进我国林业现代化发展[J].中国林业经济,2013(6):24-26.LI Genqiu. Accelerate the construction of forestry information to promote the development of China's forestry modernization[J].China Forestry Economics,2013(6):24-26.(in Chinese)
[4]刘清旺,李世明,李增元,等.无人机激光雷达与摄影测量林业应用研究进展[J].林业科学,2017,53(7):134-148.LIU Qingwang,LI Shiming,LI Zengyuan,et al. Review on the applications of UAV-based Li DAR and photogrammetry in forestry[J]. Scientia Silvae Sinicae,2017,53(7):134-148.(in Chinese)
[5]张军国,韩欢庆,胡春鹤,等.基于无人机多光谱图像的云南松虫害区域识别方法[J/OL].农业机械学报,2018,49(5):249-255.ZHANG Junguo,HAN Huanqing,HU Chunhe,et al. Identification method of Pinus yunnanensis pest area based on UAV multispectral images[J/OL]. Transactions of the Chinese Society for Agricultural Machinery,2018,49(5):249-255. http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20180529&flag=1&journal_id=jcsam. DOI:10. 6041/j.issn. 1000-1298. 2018. 05. 029.(in Chinese)
[6]李冠林,马占鸿,黄冲,等.基于K-means硬聚类算法的葡萄病害彩色图像分割方法[J].农业工程学报,2010,26(增刊2):32-37.LI Guanlin,MA Zhanhong,HUANG Chong,et al. Segmentation of color images of grape diseases using K-means clustering algorithm[J]. Transactions of the CSAE,2010,26(Supp. 2):32-37.(in Chinese)
[7]江怡,梅小明,邓敏,等.一种结合形态滤波和标记分水岭变换的遥感图像分割方法[J].地理与地理信息科学,2013,29(2):17-21.JIANG Yi,MEI Xiaoming,DENG Min,et al. Segmentation of remote sensing imagery combined with morphological filtering and marked watershed transformation[J]. Geography and Geo-Information Science,2013,29(2):17-21.(in Chinese)
[8]赵瑶池,胡祝华,白勇,等.基于纹理差异度引导的DRLSE病虫害图像精准分割方法[J/OL].农业机械学报,2015,46(2):14-19.ZHAO Yaochi,HU Zhuhua,BAI Yong,et al. An accurate segmentation approach for disease and pest based on texture difference guided DRLSE[J/OL]. Transactions of the Chinese Society for Agricultural Machinery,2015,46(2):14-19. http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20150203&flag=1&journal_id=jcsam. DOI:10. 6041/j.issn. 1000-1298. 2015. 02. 003.(in Chinese)
[9]张军国,冯文钊,胡春鹤,等.无人机航拍林业虫害图像分割复合梯度分水岭算法[J].农业工程学报,2017,33(14):93-99.ZHANG Junguo,FENG Wenzhao,HU Chunhe,et al. Image segmentation method for forestry unmanned aerial vehicle pest monitoring based on composite gradient watershed algorithm[J]. Transactions of the CSAE,2017,33(14):93-99.(in Chinese)
[10] REN S,HE K,GIRSHICK R,et al. Faster R-CNN:towards real-time object detection with region proposal networks[J].IEEE Trans. Pattern Anal. Mach. Intell.,2017,39(6):1137-1149.
[11] LIU W,ANGUELOV D,ERHAN D,et al. SSD:single shot multibox detector[C]∥European Conference on Computer Vision. Springer,Cham,2016:21-37.
[12] CHEN L C,PAPANDREOU G,KOKKINOS I,et al. Deep Lab:semantic image segmentation with deep convolutional nets,atrous convolution,and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2018,40(4):834-848.
[13] RONNEBERGER O,FISCHER P,BROX T. U-Net:convolutional networks for biomedical image segmentation[C]∥International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer,Cham,2015:234-241.
[14] HE K,ZHANG X,REN S,et al. Deep residual learning for image recognition[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2016:770-778.
[15] LONG J,SHELHAMER E,DARRELL T. Fully convolutional networks for semantic segmentation[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2015:3431-3440.
[16]孙钰,韩京冶,陈志泊,等.基于深度学习的大棚及地膜农田无人机航拍监测方法[J/OL].农业机械学报,2018,49(2):133-140.SUN Yu,HAN Jingzhi,CHEN Zhibo,et al. Monitoring method for UAV image of greenhouse and plastic-mulched landcover based on deep learning[J/OL]. Transactions of the Chinese Society for Agricultural Machinery,2018,49(2):133-140.http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20180218&flag=1&journal_id=jcsam. DOI:10.6041/j. issn. 1000-1298. 2018. 02. 018.(in Chinese)
[17] SIMONYAN K,ZISSERMAN A. Very deep convolutiona lnetworks for large-scale image recognition[C]∥International Conference on Learning Representations,2014:1-14.
[18] PAN S J,YANG Q. A survey on transfer learning[J]. IEEE Transactions on Knowledge&Data Engineering,2010,22(10):1345-1359.
[19]庄福振,罗平,何清,等.迁移学习研究进展[J].软件学报,2015,26(1):26-39.ZHUANG Fuzhen,LUO Ping,HE Qing,et al. Survey on transfer learning research[J]. Journal of Software,2015,26(1):26-39.(in Chinese)
[20] DENG J,DONG W,SOCHER R,et al. Image Net:a large-scale hierarchical image database[C]∥Computer Vision and Pattern Recognition,2009. CVPR 2009. IEEE Conference on. IEEE,2009:248-255.
[21] ZHOU B,ZHAO H,PUIG X,et al. Scene parsing through ADE20K dataset[C]∥Computer Vision and Pattern Recognition.IEEE,2017:5122-5130.
[22]张新伟,易克传,高连兴.基于脉冲耦合神经网络的粘连玉米种子图像分割[J].中国农业大学学报,2015,20(3):208-215.ZHANG Xinwei,YI Kechuan,GAO Lianxing. Image segmentation of touching corn seeds based on PCNN[J]. Journal of China Agricultural University,2015,20(3):208-215.(in Chinese)
[2]宋玉双,苏宏钧,于海英,等. 2006—2010年我国林业有害生物灾害损失评估[J].中国森林病虫,2011(6):1-4.SONG Yushuang,SU Hongjun,YU Haiying,et al. Evaluation of economic losses caused by forest pest disasters between 2006and 2010 in China[J]. Forest Pest&Disease,2011(6):1-4.(in Chinese)
[3]李根秋.加快林业信息化建设促进我国林业现代化发展[J].中国林业经济,2013(6):24-26.LI Genqiu. Accelerate the construction of forestry information to promote the development of China's forestry modernization[J].China Forestry Economics,2013(6):24-26.(in Chinese)
[4]刘清旺,李世明,李增元,等.无人机激光雷达与摄影测量林业应用研究进展[J].林业科学,2017,53(7):134-148.LIU Qingwang,LI Shiming,LI Zengyuan,et al. Review on the applications of UAV-based Li DAR and photogrammetry in forestry[J]. Scientia Silvae Sinicae,2017,53(7):134-148.(in Chinese)
[5]张军国,韩欢庆,胡春鹤,等.基于无人机多光谱图像的云南松虫害区域识别方法[J/OL].农业机械学报,2018,49(5):249-255.ZHANG Junguo,HAN Huanqing,HU Chunhe,et al. Identification method of Pinus yunnanensis pest area based on UAV multispectral images[J/OL]. Transactions of the Chinese Society for Agricultural Machinery,2018,49(5):249-255. http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20180529&flag=1&journal_id=jcsam. DOI:10. 6041/j.issn. 1000-1298. 2018. 05. 029.(in Chinese)
[6]李冠林,马占鸿,黄冲,等.基于K-means硬聚类算法的葡萄病害彩色图像分割方法[J].农业工程学报,2010,26(增刊2):32-37.LI Guanlin,MA Zhanhong,HUANG Chong,et al. Segmentation of color images of grape diseases using K-means clustering algorithm[J]. Transactions of the CSAE,2010,26(Supp. 2):32-37.(in Chinese)
[7]江怡,梅小明,邓敏,等.一种结合形态滤波和标记分水岭变换的遥感图像分割方法[J].地理与地理信息科学,2013,29(2):17-21.JIANG Yi,MEI Xiaoming,DENG Min,et al. Segmentation of remote sensing imagery combined with morphological filtering and marked watershed transformation[J]. Geography and Geo-Information Science,2013,29(2):17-21.(in Chinese)
[8]赵瑶池,胡祝华,白勇,等.基于纹理差异度引导的DRLSE病虫害图像精准分割方法[J/OL].农业机械学报,2015,46(2):14-19.ZHAO Yaochi,HU Zhuhua,BAI Yong,et al. An accurate segmentation approach for disease and pest based on texture difference guided DRLSE[J/OL]. Transactions of the Chinese Society for Agricultural Machinery,2015,46(2):14-19. http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20150203&flag=1&journal_id=jcsam. DOI:10. 6041/j.issn. 1000-1298. 2015. 02. 003.(in Chinese)
[9]张军国,冯文钊,胡春鹤,等.无人机航拍林业虫害图像分割复合梯度分水岭算法[J].农业工程学报,2017,33(14):93-99.ZHANG Junguo,FENG Wenzhao,HU Chunhe,et al. Image segmentation method for forestry unmanned aerial vehicle pest monitoring based on composite gradient watershed algorithm[J]. Transactions of the CSAE,2017,33(14):93-99.(in Chinese)
[10] REN S,HE K,GIRSHICK R,et al. Faster R-CNN:towards real-time object detection with region proposal networks[J].IEEE Trans. Pattern Anal. Mach. Intell.,2017,39(6):1137-1149.
[11] LIU W,ANGUELOV D,ERHAN D,et al. SSD:single shot multibox detector[C]∥European Conference on Computer Vision. Springer,Cham,2016:21-37.
[12] CHEN L C,PAPANDREOU G,KOKKINOS I,et al. Deep Lab:semantic image segmentation with deep convolutional nets,atrous convolution,and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2018,40(4):834-848.
[13] RONNEBERGER O,FISCHER P,BROX T. U-Net:convolutional networks for biomedical image segmentation[C]∥International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer,Cham,2015:234-241.
[14] HE K,ZHANG X,REN S,et al. Deep residual learning for image recognition[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2016:770-778.
[15] LONG J,SHELHAMER E,DARRELL T. Fully convolutional networks for semantic segmentation[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2015:3431-3440.
[16]孙钰,韩京冶,陈志泊,等.基于深度学习的大棚及地膜农田无人机航拍监测方法[J/OL].农业机械学报,2018,49(2):133-140.SUN Yu,HAN Jingzhi,CHEN Zhibo,et al. Monitoring method for UAV image of greenhouse and plastic-mulched landcover based on deep learning[J/OL]. Transactions of the Chinese Society for Agricultural Machinery,2018,49(2):133-140.http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20180218&flag=1&journal_id=jcsam. DOI:10.6041/j. issn. 1000-1298. 2018. 02. 018.(in Chinese)
[17] SIMONYAN K,ZISSERMAN A. Very deep convolutiona lnetworks for large-scale image recognition[C]∥International Conference on Learning Representations,2014:1-14.
[18] PAN S J,YANG Q. A survey on transfer learning[J]. IEEE Transactions on Knowledge&Data Engineering,2010,22(10):1345-1359.
[19]庄福振,罗平,何清,等.迁移学习研究进展[J].软件学报,2015,26(1):26-39.ZHUANG Fuzhen,LUO Ping,HE Qing,et al. Survey on transfer learning research[J]. Journal of Software,2015,26(1):26-39.(in Chinese)
[20] DENG J,DONG W,SOCHER R,et al. Image Net:a large-scale hierarchical image database[C]∥Computer Vision and Pattern Recognition,2009. CVPR 2009. IEEE Conference on. IEEE,2009:248-255.
[21] ZHOU B,ZHAO H,PUIG X,et al. Scene parsing through ADE20K dataset[C]∥Computer Vision and Pattern Recognition.IEEE,2017:5122-5130.
[22]张新伟,易克传,高连兴.基于脉冲耦合神经网络的粘连玉米种子图像分割[J].中国农业大学学报,2015,20(3):208-215.ZHANG Xinwei,YI Kechuan,GAO Lianxing. Image segmentation of touching corn seeds based on PCNN[J]. Journal of China Agricultural University,2015,20(3):208-215.(in Chinese)
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |