[1]张苏楠,田建艳,菅垄,等.基于帧间差分法-单点多框检测器的圈养生猪打斗行为识别方法[J].江苏农业学报,2021,(02):397-404.[doi:doi:10.3969/j.issn.1000-4440.2021.02.016]
　ZHANG Su-nan,TIAN Jian-yan,JIAN Long,et al.Recognition method of aggressive behaviors among pigs in pigpens based on frame difference(FD)-single shot MultiBox detector(SSD)[J].,2021,(02):397-404.[doi:doi:10.3969/j.issn.1000-4440.2021.02.016]
点击复制

基于帧间差分法-单点多框检测器的圈养生猪打斗行为识别方法()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2021年02期
页码:
397-404
栏目:
畜牧兽医·水产养殖
出版日期:
2021-04-30

文章信息/Info

Title:
Recognition method of aggressive behaviors among pigs in pigpens based on frame difference(FD)-single shot MultiBox detector(SSD)
作者:
张苏楠; 田建艳; 菅垄; 姬政雄
（太原理工大学电气与动力工程学院，山西太原030024）
Author(s):
ZHANG Su-nan; TIAN Jian-yan; JIAN Long; JI Zheng-xiong
(College of Electrical and Power Engineering, Taiyuan University of Technology, Taiyuan 030024, China)
关键词:
生猪; 打斗行为; 帧间差分法; 单点多框检测器; 判别方法
Keywords:
live pig; aggressive behaviors; frame difference; single shot MultiBox detector; discriminant method
分类号:
TP391.41
DOI:
doi:10.3969/j.issn.1000-4440.2021.02.016
文献标志码:
A
摘要:
在集约化养殖过程中，生猪打斗行为是影响生猪福利养殖的重要因素之一。针对复杂养殖环境下传统方法识别圈养生猪打斗行为准确率低的问题，提出1种基于帧间差分法（Frame difference, FD）-单点多框检测器（Single shot MultiBox detector, SSD）的生猪打斗行为识别方法。首先，利用帧间差分法提取生猪连续视频帧中的移动像素，排除光照度变化、地面水渍及尿渍等环境因素以及静止生猪对打斗行为识别的干扰。然后，以连续视频帧中的移动像素为样本，采用MobileNet_v2、焦点损失函数、网络参数迁移学习对单点多框检测器进行改进，用于检测发生剧烈运动的生猪个体，提高SSD对运动生猪个体的检测精度与速度。最后，针对生猪发生打斗行为时的特点，设计精准的生猪打斗行为判别方法，以识别生猪是否发生打斗行为。试验结果表明，该方法对生猪打斗行为的识别准确率、查准率、查全率分别达到93.75%、96.79%、90.50%，可以有效识别圈养生猪的打斗行为，为饲养员判断生猪异常状况提供依据。
Abstract:
Aggressive behavior of live pigs is one of the most important factors influencing welfare breeding of live pigs in the process of intensive breeding. Aiming at the problems of low accuracy rate of traditional methods in detecting aggressive behaviors of live pigs in pigpens under complex breeding environment, an identification method for porcine aggressive behaviors was proposed based on frame difference (FD)-single shot MultiBox detector (SSD). Firstly, FD method was used to extract the moving pixels in continuous video frames of live pigs to eliminate the interference of environmental factors such as illuminance change, surface water stain, urine stain and static pigs on the detecting of aggressive behaviors. Secondly, using the moving pixels in continuous video frames as samples, the SSD was improved by using MobileNet_v2, focal loss function and transfer learning of network parameters to detect the violently moving live pigs and improve the detection accuracy and speed of SSD. Finally, according to the characteristics of live pigs with aggressive behaviors, accurate discriminative method for live porcine aggressive behavior was designed to recognize aggressive behaviors. The experimental results showed that, the accuracy rate, precision rate and recall rate of the proposed method in recognizing porcine aggressive behaviors reached 93.75%, 96.79% and 90.50% respectively. The method can effectively recognize the aggressive behaviors of live pigs in pigpens and provide judgment basis for breeders.

参考文献/References:

[1]李丹,陈一飞,李行健,等. 计算机视觉技术在猪行为识别中应用的研究进展[J].中国农业科技导报,2019,21(7):59-69.
[2]STUKENBORG A, TRAULSEN I, PUPPE B, et al. Agonistic behaviour after mixing in pigs under commercial farm conditions[J]. Applied Animal Behaviour Science, 2011, 129(1):28-35.
[3]TURNER S P, FARNWORTH M J, WHITE I M S, et al. The accumulation of skin lesions and their use as a predictor of individual aggressiveness in pigs[J]. Applied Animal Behaviour Science, 2006, 96(3/4):245-259.
[4]STOOKEY J M, GONYOU H W. The effect of regrouping on behavioral and production parameters in finishing swine[J]. Journal of Animal Science, 1994, 72(11):2804-2811.
[5]KONGSTED A G. Stress and fear as possible mediators of reproduction problems in group housed sows: a review[J]. Acta Agriculturae Scandinavica, Section A-Animal Science, 2004, 54(2):58-66.
[6]BRACKE M B M, METZ J H M, SPRUIJT B M, et al. Decision support system for overall welfare assessment in pregnant sows B: validation by expert opinion[J]. Journal of Animal Science, 2002, 80(7):1835-1845.
[7]VIAZZI S, ISMAYILOVA G, OCZAK M, et al. Image feature extraction for classification of aggressive interactions among pigs[J]. Computers and Electronics in Agriculture, 2014, 104:57-62.
[8]OCZAK M, VIAZZI S, ISMAYILOVA G, et al. Classification of aggressive behaviour in pigs by activity index and multilayer feed forward neural network[J]. Biosystems Engineering, 2014, 119:89-97.
[9]LEE J, JIN L, PARK D, et al. Automatic recognition of aggressive behavior in pigs using a Kinect depth sensor[J]. Sensors, 2016, 16(5):631.
[10]CHEN C, ZHU W X, MA C H, et al. Image motion feature extraction for recognition of aggressive behaviors among group-housed pigs[J]. Computers and Electronics in Agriculture, 2017, 142: 380-387.
[11]CHEN C, ZHU W X, GUO Y Z, et al. A kinetic energy model based on machine vision for recognition of aggressive behaviours among group-housed pigs[J]. Livestock Science, 2018, 218:70-78.
[12]CHEN C, ZHU W X, LIU D, et al. Detection of aggressive behaviours in pigs using a RealSence depth sensor[J]. Computers and Electronics in Agriculture, 2019, 166:105003.
[13]JENSEN P, YNGVESSON J. Aggression between unacquainted pigs-sequential assessment and effects of familiarity and weight[J]. Applied Animal Behaviour Science, 1998, 58(1/2):49-61.
[14]MCGLONE J J. A quantitative ethogram of aggressive and submissive behaviors in recently regrouped pigs[J]. Journal of Animal Science, 1985, 61(3):559-565.
[15]王璐,高林,闫磊,等. 基于光流与熵统计法的花卉生长视频关键帧提取算法[J].农业工程学报,2012,28(17):125-130.
[16]李晓振,徐岩,吴作宏,等. 基于注意力神经网络的番茄叶部病害识别系统[J].江苏农业学报,2020,36(3):561-568.
[17]刘慧,张礼帅,沈跃,等. 基于改进SSD的果园行人实时检测方法[J].农业机械学报,2019,50(4):29-35,101.
[18]任胜男,孙钰,张海燕,等. 基于one-shot学习的小样本植物病害识别[J].江苏农业学报,2019,35(5):1061-1067.
[19]何东健,刘建敏,熊虹婷,等. 基于改进YOLO v3模型的挤奶奶牛个体识别方法[J].农业机械学报,2020,51(4):250-260.
[20]邓壮来,汪盼,宋雪桦,等. 基于SSD的粮仓害虫检测研究[J].计算机工程与应用,2020,56(11):214-218.
[21]FU L H, DUAN J L, ZOU X J, et al. Banana detection based on color and texture features in the natural environment[J]. Computers and Electronics in Agriculture, 2019, 167:105057.

备注/Memo

备注/Memo:
收稿日期：2020-08-27基金项目：国家高技术研究发展计划（“863”计划）项目（2013AA102306）作者简介：张苏楠（1991-），男，山西忻州人，博士研究生，主要从事复杂系统建模与智能监控研究。(Tel)18649513155;（E-mail）zhangjianjianghu05@163.com通讯作者：田建艳，（E-mail）tut_tianjy@163.com
更新日期/Last Update: 2021-05-10