期刊文献+
共找到2篇文章
< 1 >
每页显示 20 50 100
基于改进YOLO v3模型的挤奶奶牛个体识别方法 被引量:46
1
作者 何东健 刘建敏 +1 位作者 熊虹婷 芦忠忠 《农业机械学报》 EI CAS CSCD 北大核心 2020年第4期250-260,共11页
为实现无接触、高精度养殖场环境下奶牛个体的有效识别,提出了基于改进YOLO v3深度卷积神经网络的挤奶奶牛个体识别方法。首先,在奶牛进、出挤奶间的通道上方安装摄像机,定时、自动获取奶牛背部视频,并用视频帧分解技术得到牛背部图像;... 为实现无接触、高精度养殖场环境下奶牛个体的有效识别,提出了基于改进YOLO v3深度卷积神经网络的挤奶奶牛个体识别方法。首先,在奶牛进、出挤奶间的通道上方安装摄像机,定时、自动获取奶牛背部视频,并用视频帧分解技术得到牛背部图像;用双边滤波法去除图像噪声,并用像素线性变换法增强图像亮度和对比度,通过人工标注标记奶牛个体编号;为适应复杂环境下的奶牛识别,借鉴Gaussian YOLO v3算法构建了优化锚点框和改进网络结构的YOLO v3识别模型。从89头奶牛的36790幅背部图像中,随机选取22074幅为训练集,其余图像为验证集和测试集。识别结果表明,改进YOLO v3模型的识别准确率为95.91%,召回率为95.32%,m AP为95.16%,Io U为85.28%,平均帧率为32 f/s,识别准确率比YOLO v3高0.94个百分点,比Faster R-CNN高1.90个百分点,检测速度是Faster R-CNN的8倍,背部为纯黑色奶牛的F1值比YOLO v3提高了2.75个百分点。本文方法具有成本低、性能优良的特点,可用于养殖场复杂环境下挤奶奶牛个体的实时识别。 展开更多
关键词 挤奶奶牛 个体识别 视频分析 机器视觉 改进YOLO V3
下载PDF
Milk Quality in Dual Purpose Cattle with Hand or Machine Milking 被引量:1
2
作者 C. Luna-Palomera K. S. Aguilar-Hemandez J. R. Velitzquez-Martinez J. A. Peralta-Torres J.A.Aguilar-Cabrales 《Journal of Agricultural Science and Technology(A)》 2011年第8期1269-1274,共6页
To evaluate and improve the milk quality in dual purpose cattle with hand or machine milking, a good management and sanitary program was implemented (GMSP) in 40 farms. Was obtained a diagnostic by an interview and ... To evaluate and improve the milk quality in dual purpose cattle with hand or machine milking, a good management and sanitary program was implemented (GMSP) in 40 farms. Was obtained a diagnostic by an interview and in situ evaluation of management conditions and milk samples for laboratory analysis integrating the platform and microbiological counts, before and after of GMSP. The changes achieved were no significant (P 〉 0.05) for pH and density, but alcohol and foreign material were positive modified (P 〈 0.05) by GMSP in both hand and machine milking. The quantities of sediments were decreased (P 〈 0.05) in more than 50.0%. Decrease (P 〈 0.05) was observed in the counts of CFU mL1 (colony forming unitsl) in aerobic mesophilic, coliforms, Salmonella ssp. and Staphylococcus aureus with the GMSP. After GMSP Salmonella ssp. incidence was affected (P 〈 0.05) by hand (55.5%) vs. machine (61.5%) milking type; coliforms count and positive incidence in milk were affected (P 〈 0.05) in hand (81.5%) vs. machine (53.9%). Therefore it was confirmed that the implementation of adequate sanitary and management practices, in both hand and machine milking, decreased the presence and incidence of microorganisms in milk with potential to produce disease in animals and humans. 展开更多
关键词 Good livestock management practices livestock tropical systems food safety
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部