基于改进YOLOX的群养生猪轻量化目标检测方法

邓铭辉; 龚俊杰; 郑飘逸; 马闯; 尹艳玲

基于改进YOLOX的群养生猪轻量化目标检测方法

东北农业大学电气与信息学院

基金项目:

国家自然科学基金面上项目(32172784)

详细信息

作者简介:
邓铭辉(1976—),男,副教授,博士,主要从事农业人工智能研究,E-mail:markdmh@163.com

中图分类号: TP183;TP391.41;S828
计量
- 文章访问数: 1
- HTML全文浏览量: 0
- PDF下载量: 0
出版历程
- 收稿日期: 2023-05-07
- 刊出日期: 2023-11-24

Lightweight Target Detection Method for Group-raised Pigs Based on Improved YOLOX

College of Electrical Engineering and Information, Northeast Agricultural University

摘要

摘要: 针对目前群养生猪智能化养殖中复杂环境下猪只目标检测精度低的问题，提出了一种基于改进YOLOX的群养生猪轻量化目标检测模型Ghost-YOLOX-BiFPN。该模型采用Ghost卷积替换普通卷积，在减少主干网络参数的情况下，提高了模型的特征提取能力。使用加入CBAM注意力机制的BiFPN作为模型的Neck部分，使得模型充分融合不同体型猪只的特征图，并使用Focal Loss损失函数解决猪圈环境下猪只与背景难以区分的问题，增强模型对正样本的学习。实验结果表明，改进后模型对群养生猪检测精度为95.80%,相比于原始YOLOX算法，检测精度提升2.84个百分点，参数量降低63%。最后将本文轻量化模型部署到Nvidia Jetson Nano移动端开发板，通过在开发板上实际运行表明，本文所提模型实现了对不同大小、不同品种猪只的准确识别，为后续智能化生猪养殖提供支持。
- 猪只目标检测 /
- YOLOX /
- 图像处理 /
- 移动端部署
Abstract: Aiming at the problem of low pig target detection accuracy in the complex environment in the current intelligent breeding of group-raised pigs, a lightweight target detection model for group-raised pigs based on improved YOLOX, Ghost-YOLOX-BiFPN was proposed. The Ghost convolution was used to replace the traditional convolution, which greatly reduced the number of model parameters. BiFPN was used as the model feature fusion network to effectively fuse the feature maps of pigs of different sizes, and Focal Loss function was added in the post-processing stage, increasing the learning of the model to the positive sample target, and reducing the rate of missed detection. The results showed that the improved model had a detection accuracy of 95.80% for pigs, and the number of model parameters were 2.001×10~7. Compared with the original YOLOX algorithm, the detection accuracy and recall were increased by 2.84 percentage points and 3.22 percentage points, respectively, and the number of model parameters were reduced by 63%. Finally, the proposed algorithm model was deployed to the Nvidia Jetson Nano mobile terminal development board. The actual operation on the development board showed that the model proposed can guarantee the recognition rate of pigs and realize the accurate recognition of pigs of different sizes and breeds. The research result can provide support for the subsequent establishment of intelligent pig breeding system.
- pig target detection /
- YOLOX /
- image processing /
- mobile deployment

HTML全文

参考文献(26)

[1]	孙东来,王继超,陈科,等.基于Ghost-YOLOv3-2算法的2尺度猪目标检测[J].江苏农业科学,2022,50(7):189-196.SUN Donglai,WANG Jichao,CHEN Ke,et al.2-scale pig target detection based on Ghost-YOLOv3-2 algorithm[J].Jiangsu Agricultural Science,2022,50(7):189-196.(in Chinese)
[2]	SAKIR T,URKMEZ A,INAL S.Determination of body measurements on the Holstein cows using digital image analysis and estimation of live weight with regression analysis[J].Computers and Electronics in Agriculture,2011,76(2):189-197.
[3]	SAKIR T,URKMEZ A,INAL S.A fuzzy rule-based system for predicting the live weight of Holstein cows whose body dimensions were determined by image analysis[J].Turkish Journal of Electrical Engineering and Computer Sciences,2011,19(4):689-703.
[4]	OCHOA Z,LUIS E.Evaluation of the accuracy of simple body measurements for live weight prediction in growing- finishing pigs[D].Champaign:University of Illinois at Urbana-Champaign,2010.
[5]	ENEVOLDSEN C,KRISTENSEN T.Estimation of body weight from body size measurements and body condition scores in dairy cows[J].Journal of Dairy Science,1997,80(9):1988-1995.
[6]	李菊霞,李艳文,牛帆,等.基于YOLO v4的猪只饮食行为检测方法[J].农业机械学报,2021,52(3):251-256.LI Juxia,LI Yanwen,NIU Fan,et al.Pig diet behavior detection method based on YOLO v4[J].Transactions of the Chinese Society for Agricultural Machinery,2021,52(3):251-256.(in Chinese)
[7]	杨心,朱伟兴.基于广义Hough聚类的粘连猪头尾识别[J].江苏农业科学,2018,46(9):230-235.YANG Xin,ZHU Weixing.Recognition of adhesive pig head and tail based on generalized Hough clustering[J].Jiangsu Agricultural Science,2018,46(9):230-235.(in Chinese)
[8]	段玉瑶,马丽,刘刚.猪舍场景下的生猪目标跟踪和行为检测方法研究[J].农业机械学报,2015,46(增刊):187-193.DUAN Yuyao,MA Li,LIU Gang.Target tracking and behavior detection method in piggery scenarios[J].Transactions of the Chinese Society for Agricultural Machinery,2015,46(Supp.):187-193.(in Chinese)
[9]	马丽,纪滨,刘宏申,等.单只猪轮廓图的侧视图识别[J].农业工程学报,2013,29(10):168-174.MA Li,JI Bin,LIU Hongshen,et al.Side view recognition of single pig outline[J].Transactions of the CSAE,2013,29(10):168-174.(in Chinese)
[10]	谢双云.基于图像序列的猪目标检测和跟踪研究[D].太原:太原理工大学,2015.XIE Shuangyun.Research on pig target detection and tracking based on image sequence[D].Taiyuan:Taiyuan University of Technology,2015.(in Chinese)
[11]	朱家骥,张立生.基于图像处理的单只猪目标检测方法研究[J].信息与电脑(理论版),2014(24):11-12.ZHU Jiaji,ZHANG Lisheng.Research on single pig target detection method based on image processing[J].Information and Computer (Theoretical Edition),2014(24):11-12.(in Chinese)
[12]	宁远霖,杨颖,李振波,等.基于改进YOLOv5的复杂跨域场景下的猪个体识别与计数[J].农业工程学报,2022,38(17):168-175.NING Yuanlin,YANG Ying,LI Zhenbo,et al.Pig individual recognition and counting in complex cross-domain scenarios based on improved YOLOv5[J].Transactions of the CSAE,2022,38(17):168-175.(in Chinese)
[13]	燕红文,刘振宇,崔清亮,等.基于特征金字塔注意力与深度卷积网络的多目标生猪检测[J].农业工程学报,2020,36(11):193-202.YAN Hongwen,LIU Zhenyu,CUI Qingliang,et al.Multi-target pig detection based on feature pyramid attention and deep convolutional network[J].Transactions of the CSAE,2020,36(11):193-202.(in Chinese)
[14]	沈明霞,太猛,CEDRIC Okinda,等.基于深层卷积神经网络的初生仔猪目标实时检测方法[J].农业机械学报,2019,50(8):270-279.SHEN Mingxia,TAI Meng,CEDRIC Okinda,et al.Real-time detection method of newborn piglets based on deep convolution neural network[J].Transactions of the Chinese Society for Agricultural Machinery,2019,50(8):270-279.(in Chinese)
[15]	ZHAO Hengshuang,SHI Jianping,QI Xiaojuan,et al.Pyramid scene parsing network[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR),2017:2881-2890.
[16]	KIRILLOV A,GIRSHICK R,HE K,et al.Panoptic feature pyramid networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR),2019:6399-6408.
[17]	CHEN Liangjie,ZHU Yukun,PAPANDREOU G,et al.Encoder-decoder with atrous separable convolution for semantic image segmentation[C]//Proceedings of the European Conference on Computer Vision (ECCV),2018:801-818.
[18]	GE Z,LIU S,WANG F,et al.YOLOX:exceeding YOLO series in 2021[J].ArXiv preprint,arXiv:2107.08430,2021.
[19]	LIU S,QI L,QIN H,et al.Path aggregation network for instance segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2018:8759-8768.
[20]	WOO S,PARK J,LEE J Y,et al.CBAM:convolutional block attention module[C]//Proceedings of the 2018 European Conference on Computer Vision,2018:3-19.
[21]	TAN M,PANG R,LE Q V.EfficientDet:scalable and efficient object detection[C]//IEEE Conference on Computer Vision and Pattern Recognition,2020:10778-10787.
[22]	MAO Keyang,KUN Yu,ZHANG Chi,et al.Dense aspp for semantic segmentation in street scenes[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR),2018:3684-3692.
[23]	WANG Fei,JIANG Mengqing,QIAN Chen,et al.Residual attention network for image classification[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR),2017:3156-3164.
[24]	YU Changqian,GAO Changxin,WANG Jingbo,et al.BiSeNet:bilateral segmentation network for real-time semantic segmentation[J].International Journal of Computer Vision,2021,129(11):3051-3068.
[25]	ZHAO Ting,WU Xiangqian.Pyramid feature attention network for saliency detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR),2019:3085-3094.
[26]	LIN T Y,GOYAL P,GIRSHICK R.Focal loss for dense object detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2020,42(2):318-327.