A total of 60 crossbred(Large White×Landrace) pigs of halothane genotype NN(castrated males and females) were allotted to three treatments:3 h lairage with toys,3 h lairage and 0 h lairage in a randomized complet...A total of 60 crossbred(Large White×Landrace) pigs of halothane genotype NN(castrated males and females) were allotted to three treatments:3 h lairage with toys,3 h lairage and 0 h lairage in a randomized complete block design and used to evaluate the influence of lairage conditions on behavior, biochemical indicators and meat quality for finishing pigs at slaughter.Behavior of the pigs was scored subjectively during lairage.Blood samples were taken at exsanguination s to measure blood temperature, plasma Cortisol,ACTH,glucose,lactate,plasma enzymes and hematological indices.Post-mortem meat quality measurements included muscle colour value(MCV),electrical conductivity(EC),pH at 45 min and 24 h from Longissimus thoracis(LM) and Semimembranosus(SM) mucles and drip loss from LM. The results showed that 3 h lairage group with toys demonstrated significantly improved behavior than the group without toys at 3 sampling times.All the pigs showed increasing calmness as the time of lairage progressed.The omission of lairage increased plasma Cortisol,ACTH,glucose and lactate(P【0.05),and decreased plasma lactate dehydrogenase(LDH),and creatine kinase(CK)(P【0.05).No biochemical index was influenced by the presence or absence of toys during lairage(P【0.05).Muscle colour value, electrical conductivity,pH at 45 min and 24 h from LM and SM and drip loss were not affected by any treatment(P【0.05).Pigs provided 3 h lairage,with or without toys,exhibited lower red blood cell(RBC), hemoglobin(HGB),and haematocrit(HCT) when compared to 0 h lairage.3 h lairage with or without toys resulted in higher white blood cell(WBC) and lymphocyte(W-SCC) levels than 0 h lairage.None of the hemocytic indices in pigs given lairage was affected by the presence or absence of toys.We conclude from this pilot study that in local commercial conditions,from the point of view of animal welfare and meat quality,lairage time of 3 h after short travel was beneficial.Pigs resting showed increased relief from stress and a recovery in immune competence.Holding pigs in lairage with toys for a few hours after arrival at the abattoir may be beneficial for the animal’s well-being.展开更多
基于视频的生猪行为跟踪和识别对于实现精细化养殖具有重要价值。为了应对群养生猪多目标跟踪任务中由猪只外观相似、遮挡交互等因素带来的挑战,研究提出了基于PigsTrack跟踪器的群养生猪多目标跟踪方法。PigsTrack跟踪器利用高性能YOLO...基于视频的生猪行为跟踪和识别对于实现精细化养殖具有重要价值。为了应对群养生猪多目标跟踪任务中由猪只外观相似、遮挡交互等因素带来的挑战,研究提出了基于PigsTrack跟踪器的群养生猪多目标跟踪方法。PigsTrack跟踪器利用高性能YOLOX网络降低目标误检与漏检率,采用Transformer模型获取具有良好区分特性的目标外观特征;基于OC-SORT(observation-centric sort)的思想,通过集成特征匹配、IoU匹配和遮挡恢复匹配策略实现群养生猪的准确跟踪。基于PBVD(pigs behaviours video dataset)数据集的试验结果表明,PigsTrack跟踪器的HOTA(higher order tracking accuracy),MOTA(multiple object tracking accuracy)和IDF1得分(identification F1 score)分别为85.66%、98.59%和99.57%,相较于现有算法的最高精度,分别提高了3.71、0.03和2.05个百分点,证明了PigsTrack跟踪器在解决外观相似和遮挡交互引起的跟踪过程中身份跳变问题方面的有效性。随后,利用Slowfast网络对PigsTrack跟踪器的跟踪结果进行了典型行为统计,结果显示PigsTrack在群养生猪个体行为统计方面更准确。此外,通过在ABVD(aggressive-behavior video)数据集上的试验,PigsTrack跟踪器的HOTA、MOTA和IDF1得分分别为69.14%、94.82%和90.11%,相对于现有算法的最高精度,提高了5.33、0.57和8.60个百分点,验证了PigsTrack跟踪器在群养生猪跟踪任务中的有效性。总而言之,PigsTrack跟踪器能够有效应对外观相似和遮挡交互等挑战,实现了准确的生猪多目标跟踪,并在行为统计方面展现出更高的准确性,为生猪养殖领域的研究和实际应用提供了有价值的指导。展开更多
针对传统的群养猪行为观察方法的缺点,提出了1种疑似病猪行为自动监测系统。系统基于ARM平台,利用安装于猪舍排泄区的嵌入式监控设备对群养猪的排泄行为进行24h监控,通过1种改进的运动目标检测算法和基于像素块对称特征的图像识别算法...针对传统的群养猪行为观察方法的缺点,提出了1种疑似病猪行为自动监测系统。系统基于ARM平台,利用安装于猪舍排泄区的嵌入式监控设备对群养猪的排泄行为进行24h监控,通过1种改进的运动目标检测算法和基于像素块对称特征的图像识别算法定位具有异常行为的疑似病猪,并将报警图像通过通用分组无线服务(general packet radio service)网络传送至监控中心。对一栏10头大约克夏猪的试验结果表明,病猪检测正确率为78.38%,基本达到了预期的目标。因此,该文设计的方法对我国的养殖业实施自动化监测具有一定的借鉴意义。展开更多
在利用视频监控技术对群养猪只进行自动行为监测时,对猪只准确定位并辨别其头尾位置对提高监测水平至关重要,基于此提出一种基于YOLO v3(You only look once v3)模型与图结构模型(Pictorial structure models)的猪只头尾辨别方法。首先...在利用视频监控技术对群养猪只进行自动行为监测时,对猪只准确定位并辨别其头尾位置对提高监测水平至关重要,基于此提出一种基于YOLO v3(You only look once v3)模型与图结构模型(Pictorial structure models)的猪只头尾辨别方法。首先,利用基于深度卷积神经网络的YOLO v3目标检测模型,训练猪只整体及其头部和尾部3类目标的检测器,从而在输入图像中获得猪只整体及头尾部所有的检测结果;然后,引入图结构模型,描述猪只的头尾结构特征,对每个猪只整体检测矩形框内的头尾部位组合计算匹配得分,选择最优的部位组合方式;对部分部位漏检的情况,采取阈值分割与前景椭圆拟合的方法,根据椭圆长轴推理出缺失部位。在实际猪场环境下,通过俯拍获得猪舍监控视频,建立了图像数据集,并进行了检测实验。实验结果表明,与直接利用YOLO v3模型相比,本文方法对头尾定位的精确率和召回率均有一定提高。本文方法对猪只头尾辨别精确率达到96.22%,与其他方法相比具有明显优势。展开更多
基金supported financially by the project‘Research and Development on Technology and Key Equipment for New Type of Industrialized and Healthy Animal Husbandry'funded by the Chinese Ministry of Science and Technology(2006BAD 14B02-6)11th Five Years Key Programs for Science and Technology Development of China(2006BAD01A08-07)+1 种基金Hubei Province Key Project of Science and Technology (2006AA201B24)Wuhan City Key Project of Industrialization(200720112026)
文摘A total of 60 crossbred(Large White×Landrace) pigs of halothane genotype NN(castrated males and females) were allotted to three treatments:3 h lairage with toys,3 h lairage and 0 h lairage in a randomized complete block design and used to evaluate the influence of lairage conditions on behavior, biochemical indicators and meat quality for finishing pigs at slaughter.Behavior of the pigs was scored subjectively during lairage.Blood samples were taken at exsanguination s to measure blood temperature, plasma Cortisol,ACTH,glucose,lactate,plasma enzymes and hematological indices.Post-mortem meat quality measurements included muscle colour value(MCV),electrical conductivity(EC),pH at 45 min and 24 h from Longissimus thoracis(LM) and Semimembranosus(SM) mucles and drip loss from LM. The results showed that 3 h lairage group with toys demonstrated significantly improved behavior than the group without toys at 3 sampling times.All the pigs showed increasing calmness as the time of lairage progressed.The omission of lairage increased plasma Cortisol,ACTH,glucose and lactate(P【0.05),and decreased plasma lactate dehydrogenase(LDH),and creatine kinase(CK)(P【0.05).No biochemical index was influenced by the presence or absence of toys during lairage(P【0.05).Muscle colour value, electrical conductivity,pH at 45 min and 24 h from LM and SM and drip loss were not affected by any treatment(P【0.05).Pigs provided 3 h lairage,with or without toys,exhibited lower red blood cell(RBC), hemoglobin(HGB),and haematocrit(HCT) when compared to 0 h lairage.3 h lairage with or without toys resulted in higher white blood cell(WBC) and lymphocyte(W-SCC) levels than 0 h lairage.None of the hemocytic indices in pigs given lairage was affected by the presence or absence of toys.We conclude from this pilot study that in local commercial conditions,from the point of view of animal welfare and meat quality,lairage time of 3 h after short travel was beneficial.Pigs resting showed increased relief from stress and a recovery in immune competence.Holding pigs in lairage with toys for a few hours after arrival at the abattoir may be beneficial for the animal’s well-being.
文摘基于视频的生猪行为跟踪和识别对于实现精细化养殖具有重要价值。为了应对群养生猪多目标跟踪任务中由猪只外观相似、遮挡交互等因素带来的挑战,研究提出了基于PigsTrack跟踪器的群养生猪多目标跟踪方法。PigsTrack跟踪器利用高性能YOLOX网络降低目标误检与漏检率,采用Transformer模型获取具有良好区分特性的目标外观特征;基于OC-SORT(observation-centric sort)的思想,通过集成特征匹配、IoU匹配和遮挡恢复匹配策略实现群养生猪的准确跟踪。基于PBVD(pigs behaviours video dataset)数据集的试验结果表明,PigsTrack跟踪器的HOTA(higher order tracking accuracy),MOTA(multiple object tracking accuracy)和IDF1得分(identification F1 score)分别为85.66%、98.59%和99.57%,相较于现有算法的最高精度,分别提高了3.71、0.03和2.05个百分点,证明了PigsTrack跟踪器在解决外观相似和遮挡交互引起的跟踪过程中身份跳变问题方面的有效性。随后,利用Slowfast网络对PigsTrack跟踪器的跟踪结果进行了典型行为统计,结果显示PigsTrack在群养生猪个体行为统计方面更准确。此外,通过在ABVD(aggressive-behavior video)数据集上的试验,PigsTrack跟踪器的HOTA、MOTA和IDF1得分分别为69.14%、94.82%和90.11%,相对于现有算法的最高精度,提高了5.33、0.57和8.60个百分点,验证了PigsTrack跟踪器在群养生猪跟踪任务中的有效性。总而言之,PigsTrack跟踪器能够有效应对外观相似和遮挡交互等挑战,实现了准确的生猪多目标跟踪,并在行为统计方面展现出更高的准确性,为生猪养殖领域的研究和实际应用提供了有价值的指导。
文摘针对传统的群养猪行为观察方法的缺点,提出了1种疑似病猪行为自动监测系统。系统基于ARM平台,利用安装于猪舍排泄区的嵌入式监控设备对群养猪的排泄行为进行24h监控,通过1种改进的运动目标检测算法和基于像素块对称特征的图像识别算法定位具有异常行为的疑似病猪,并将报警图像通过通用分组无线服务(general packet radio service)网络传送至监控中心。对一栏10头大约克夏猪的试验结果表明,病猪检测正确率为78.38%,基本达到了预期的目标。因此,该文设计的方法对我国的养殖业实施自动化监测具有一定的借鉴意义。
文摘在利用视频监控技术对群养猪只进行自动行为监测时,对猪只准确定位并辨别其头尾位置对提高监测水平至关重要,基于此提出一种基于YOLO v3(You only look once v3)模型与图结构模型(Pictorial structure models)的猪只头尾辨别方法。首先,利用基于深度卷积神经网络的YOLO v3目标检测模型,训练猪只整体及其头部和尾部3类目标的检测器,从而在输入图像中获得猪只整体及头尾部所有的检测结果;然后,引入图结构模型,描述猪只的头尾结构特征,对每个猪只整体检测矩形框内的头尾部位组合计算匹配得分,选择最优的部位组合方式;对部分部位漏检的情况,采取阈值分割与前景椭圆拟合的方法,根据椭圆长轴推理出缺失部位。在实际猪场环境下,通过俯拍获得猪舍监控视频,建立了图像数据集,并进行了检测实验。实验结果表明,与直接利用YOLO v3模型相比,本文方法对头尾定位的精确率和召回率均有一定提高。本文方法对猪只头尾辨别精确率达到96.22%,与其他方法相比具有明显优势。
