计算机视觉技术在智能养猪中的研究进展

计算机视觉是人工智能的重要发展方向,可以显著改善人与世界的交互方式,加强计算机视觉技术在智能养猪中的研究和应用必将为养猪业带来一场变革。本文综述了计算机视觉技术在猪只识别、行为监测、体尺测量、体重估测、肉品质评定、疾病监测等方面的研究进展,以期促进我国养猪业的智能化发展。     

最小二乘法在黄牛体重估测中的应用

本文能过最小二乘法用黄牛的体长、胸围两项体尺估测黄牛的体重。不同的估测方法估测的精度不同,采用体长和胸围两项体尺建立在二元二次回归方程估测的体重Ｙ１与实测的体重Ｙ０的相关系数最高（ｒ０１为０．６７０５）;用体长Ｘ１的单项体尺估测的体重Ｙ２与实际体重Ｙ０相关系数（ｒ０２）为０．６１４;用胸围Ｘ２的单项体尺估测的体重Ｙ３与实测体重Ｙ０的相关系数（ｒ０３）为０．６２６;用体长Ｘ２建立二元一次回归方程估测     

13月龄母牦牛体重与体尺指标的相关回归分析

为了分析大通种牛场牦牛体尺与体重的相关性,估测体重的回归分析,对大通牦牛母牛进行称重和体尺测量,根据得出的数据分析其体重与体高、背长、胸围的相关系数,同时估测大通牦牛母牛的体重回归模型。结果表明,大通牦牛母牛体重与体高、背长、胸围之间极显著相关,说明体重与体尺之间存在着明显的线性关系。     

大通牦牛体重与体尺指标的相关回归分析

为了分析大通牦牛体尺与体重的相关性,以便在实际工作中得到应用,对110头6~10月龄大通牦牛母牛进行称重和体尺测量,根据得出的数据分析其体重与体高、体斜长、胸围的相关系数,同时估测大通牦牛的体重回归模型。结果显示,大通牦牛体重与体高、体斜长、胸围的相关系数分别为0.702、0.879和0.977,得到了2个估测体重的回归模型,估测值与实测值之间相关程度分别为0.993、0.954、0.992和0.993。大通牦牛体重与体高、体斜长、胸围之间显著相关,说明体重与体尺之间存在着明显的线性关系。     

高原型藏系羊初生体重与体尺指标的相关性研究

为了分析初生高原型藏系羊羔羊的体尺性状对体重的影响,本试验随机选取初生12 h以内的藏系羊羔羊为研究对象,测量羔羊的体斜长、体高、胸围3个体尺数据,并称重,采用逐步线性回归的方法建立羔羊体重与体尺的多元线性回归方程。结果显示:高原型藏系羊初生羔羊的体重、体高、体长、胸围的变异系数为14.88%、8.50%、6.88%、9.85%;体重与体高、体长、胸围的表型相关系数分别为0.577、0.743、0.911,经检验差异达到极显著水平(P0.01);根据体尺性状建立三个估测体重的回归模型,得出估测值与实测值间的相关程度分别为0.842、0.827、0.833。试验表明,高原型藏系羊羔羊的体重与体高、体长、胸围之间显著相关,且体重与体尺存在着明显的线性关系。     

1.5岁大通牦牛公牛体重与体尺指标的相关回归分析

为了分析大通牦牛1.5岁公牦牛体尺与体重的相关性,以便在实际工作中应用于体重估测,在2011年11月对31头1.5岁大通牦牛公牛进行称重和体尺测量,根据得出的数据分析体重与体高、体斜长、胸围、管围的相关系数,同时建立估测大通牦牛1.5岁公牛的体重回归模型。结果表明,大通牦牛1.5岁公牛体重与体高、体斜长、胸围、管围的相关系数分别为0.603、0.579、0.453、0.600,经检验,体高、管围与体重的相关系数达到了极显著水平(P<0.01),体斜长、胸围与体重的相关系数达到了显著水平(P<0.05),得到了3个估测体重的回归模型,估测值与实测值之间相关程度分别为0.820、0.603、0.786。大通牦牛1.5岁公牛体重与体高、体斜长、胸围、管围之间显著相关,说明体重与体尺之间存在着明显的线性关系。     

2.5岁大通牦牛母牛体重与体尺指标的相关回归分析

为了分析2.5岁大通牦牛母牛体尺与体重的相关性,以便在实际工作中应用,在11月份对31头2.5岁大通牦牛母牛进行称重和体尺测量,根据得出的数据分析其体重与体高、体斜长、胸围、管围的相关系数,同时估测大通牦牛2.5岁母牛的体重回归模型。结果显示,大通牦牛2.5岁母牛体重与体高、体斜长、胸围、管围的相关系数分别为0.309、0.713、0.846和0.595,经检验体高、管围与体重的相关系数达到了显著水平(p0.05),体斜长、胸围与体重的相关系数达到了极显著水平(p0.01),得到了3个估测体重的回归模型,估测值与实测值之间相关程度分别为0.923、0.864和0.893。大通牦牛2.5岁母牛体重与体高、体斜长、胸围、管围之间显著相关,说明体重与体尺之间存在着明显的线性关系。     

大额牛体重估测研究

大额牛是我国特有的一种珍贵畜种,生活在云南省海拔2 500～3 600 m的地带.研究选用体斜长和胸围2项体尺指标,建立不同的回归方程估测大额牛体重,并对估测体重和实测体重进行了相关性分析,以期筛选出最佳体重估测方法用于指导生产实践.     

应用逐步回归法估测中国西门塔尔牛体重的研究

本研究根据内蒙古乌拉盖地区某牛场牛的体长、体高、胸围、腹围、管围等体尺指数,应用逐步回归法,对牛的体重进行估测,并比较实测体重与估测体重之间的差异。得出估测中国西门塔尔牛体重的一般公式,按此公式所作的F和t检验均表明预测值与实际体重测定值无显著差异,其相关系数值达到0.952。     

南阳牛主要性状遗传参数的初步估测

<正> 南阳牛是我国优秀的地方良种之一,为了弄清其遗传性能,我们对其主要体尺、体重的遗传力、遗传相关进行了估测。结果如下。 一、材料与方法 材料来自河南省南阳黄牛研究所1979—1983年度牛只生长发育记录,共1019个样本(8头公牛的女儿),群内无近交个体。测算了1岁半牛的体尺、体重的遗传力、遗传相关及初生重、8月龄体重、2岁半体重的遗传力。样本数在100个以上的剔除     

舍饲小尾寒羊部分生产性能的测定

为了保护生态环境,多数的草场被禁牧,舍饲养羊替代放牧饲养已成为养羊业发展的必然趋势。吉林省石油集团农工商会企业总公司农业总场小尾寒羊养殖场采用日光温室式羊舍,对种羊、怀孕羊,哺乳羊、羔羊、育肥羊,采用全年均衡配种,均衡产羔、均衡上市的生产模式,3年生产试验表明,小尾寒羊是适于舍饲的品种,舍饲养羊各项生产指标较高,经济效益较好。     

甘肃省羊小反刍兽疫发生趋势分析

自去年12月以来,小反刍兽疫在国内多地出现,严重影响了养羊业的健康发展.本文根据甘肃省养羊业现状和羊病情况,分析了小反刍兽疫的发生趋势,旨在指导行业或养殖场户认真做好羊病防控,减少损失,保障发展.     

基于物联网的智能母猪养殖管理系统

该设计通过无线通信技术为母猪养殖场设计了一款养殖信息采集装置,以达到远程控制养殖场的目的。以STM32F1单片机为中心控制器,选用红外传感器模块、空气温湿度传感器HMP45D模块,进行养殖场的母猪身理信息采样,采用GPRS无线传输模块完成数据的发送与接收,然后将接收到的数据实时处理后传送到上位机,上位机与电脑的连接采用Labview软件,Labview再与My Sql数据库连接,最终数据可以在系统中显示,实现了远程监测和数据化显示。此系统实现了母猪养殖的信息化数据化,具有推广前景。     

2016—2019年云南省昆明市小反刍兽疫集中监测

为切实做好云南省小反刍兽疫消灭工作,保障羊产业稳定健康发展,依据《国家中长期动物疫病防治规划(2012—2020年)》和《全国小反刍兽疫消灭计划(2016—2020年)》要求,2016—2019年分别采用ELISA、RT-PCR方法,在昆明市开展了小反刍兽疫免疫抗体和病毒核酸检测。结果显示:909个集中监测养殖场点的8972份血清样品中,检出免疫合格场点841个,平均场点免疫合格率为92.5%;检出免疫合格样品7405份,平均个体免疫合格率为82.53%,虽然超过了农业农村部要求的不低于70%的标准,但各年度的个体合格率有下降趋势;规模场和散养户的个体免疫合格率分别为84.97%和82.07%,场点合格率分别为83.33%和93.38%,均无显著差异(P>0.05);不同县(市、区)的个体免疫合格率为77.36~93.17%,场点合格率为83.33%~98.55%,差异均不显著(P>0.05);在3812份羊眼(鼻)棉拭子和组织样品中检出疑似核酸阳性2份,分别来自2个散养户。结果表明:昆明市小反刍兽疫整体防控效果较好,发生大面积流行的风险较低,但仍有散发疫情风险。因此,仍需继续做好小反刍兽疫的强制免疫、监测和流行病学调查工作,同时加强羊只的引种检疫和流通监管,最终到达消灭小反刍兽疫的目标。     

A computer model to simulate control of parasitic gastroenteritis in sheep on UK farms

A computer model that simulates the population dynamics and epidemiology of three major species of parasitic nematodes of sheep found in the UK (Telodorsagia [Ostertagia] spp., Haemonchus spp. and Trichostrongylus spp.) is described. The model has been developed as a tool for veterinarians and advisors to aid in the implementation of integrated parasite control strategies designed to optimise anthelmintic usage and delay the development of resistance on UK farms. The model represents the parasite life cycle, flock dynamics and the response of individuals with different susceptible and resistant genotypes to the major broad-spectrum classes of anthelmintic available in the UK. Where possible, UK data have been used for the model parameters. The model allows worm control simulations on individual UK farms. Inputs include environmental and farm management variables which impact on the epidemiology of the disease, e.g. regional weather data; flock stocking rates; initial pasture larval contamination levels and species proportions; lambing dates; timing of flock movements to clean pastures; and removal of lambs during the year. Farm management data, as well as nematode egg outputs and grass larval counts, were collected from eight UK farms over a 1-year period for initial validation of the model outputs. The management data for each farm were used as inputs for each model run and model outputs for nematode egg counts from ewes and lambs were compared to the observed data for each farm. Statistical analysis of results shows a positive correlation for observed and simulated counts and regression analysis suggests an acceptable fit between the data. Comparison of observed and simulated outputs for resistance were possible for only one farm due to low numbers of worms developing in the laboratory tests. Additional studies will be necessary before resistance data can be reliably compared. Further validation studies are proposed to ensure that the model is robust and applicable across a diverse range of farm types. The model will be used to demonstrate the advantage, in terms of delaying resistance development, of current guidelines for anthelmintic use and management practices for worm control in sheep.     

Breeding for resistance to mastitis in United Kingdom sheep, a review and economic appraisal

Mastitis is a problem in the sheep industry, and its incidence varies widely with how it is recorded, the breed of sheep and the farm. Virtually all the published information about the genetics of mastitis refers to dairy breeds of cattle and sheep, and there is little information for meat sheep breeds. Many dairy breeding programmes worldwide use the somatic cell count (scc) in milk as an indicator of resistance to clinical and subclinical mastitis, but it is difficult to measure in meat sheep breeds. Molecular genetic technologies may therefore be a more practical way to assess susceptibility to mastitis. This paper reviews the genetics of mastitis and considers the opportunities for breeding for resistance to mastitis, with particular reference to sheep. In addition, to investigate the potential economic effects of mastitis in a purebred sheep population, a computer model of flock dynamics was developed. By making a modest set of assumptions about the key farm parameters that influence lowland sheep production, the model showed that breeding for resistance (or other control methods), if it could reduce the risk of contracting mastitis by 10 per cent, would be worth pounds 8.40 per ewe, equivalent annually to pounds 2.7 million for the purebred sector of the Texel breed alone.     

内蒙古巴林左旗查干哈达苏木牲畜头数与草地面积变迁分析

为了能够给进一步计算内蒙古赤峰市巴林左旗查干哈达苏木草地载畜量时提供数据基础,同时也为了给深入研究该地区近67年来草原畜牧业发展情况时提供参考依据,在搜集了1949—2016年该地区历年牲畜头数和草地面积数据的基础上,分析了该地区近67年来牲畜头数与畜均占有草地面积变迁情况。结果表明,该地区牲畜头数从1949年的5 733羊单位增加到2016年201 326羊单位,67年间增加了35倍;畜均（羊单位）占有草地面积从5.50 hm^2/羊单位下降到0.15 hm^2/羊单位,67年间畜均占有草地面积下降了97.3％。综合分析可知,该地区草地面积逐年缩小,草地类型趋于旱生化,而牲畜头数则逐年增加,使草地压力日趋严重。     

Ultrasonographic examinations of fetal growth of sheep between day 20 and day 50 of gestation

Previous studies on embryonic and fetal growth in sheep were mostly transversal using animals killed at various stages of gestation. Until now it was difficult to monitor the development of individual embryos/foetuses during pregnancy, especially during the first and second pregnancy month. Real-time ultrasound as a non invasive method could be an appropriate method for examination of embryonic and early foetal development in sheep. The aim of this study was to determine the embryonic and foetal development of the crown-rump-length (CRL) in pregnant ewes in relation to the number of fetuses and/or the breed. Between the 20th and 50th day of pregnancy the embryos/foetuses showed an exponential growth which can be best described by the equation of the form CRL (mm) = W * exp (k * day of pregnancy). The individual variability in embryofetal growth is in part due to the number of embryos per sheep and the sheep breed.