非接触式测温鉴定母猪典型生理状态

体温是衡量母猪发情与否的关键生理指标。母猪发情期与间情期体温的明显不同，是运用红外热成像监测母猪体温变化以鉴定母猪发情的依据。该研究基于Y3TB01体温筛查智能摄像机和红外热像仪C3实时非接触拍摄母猪红外热图像，并运用图像分析技术获取体温，为非接触式测温用于鉴定母猪典型生理状态提供支持。试验以空怀期、发情期、妊娠1～8 d和妊娠9～16 d大白初产和经产母猪共720头为研究对象，利用红外设备和电子体温计分别测量母猪的眼睛、耳、耳蜗、乳房、外阴、臀部和直肠温度，筛选不同胎次发情母猪和返情母猪的体表关键部位温度，进行差异分析。结果表明：运用Y3TB01体温筛查智能摄像机可以代替直肠温度测定，准确监测母猪体表温度（相关系数为0.973）；红外热像仪C3检测体温异常母猪的体表温度，臀部温度可以作为筛选大白初产发情母猪和返情母猪的测定部位（准确率分别为77%和72%），外阴温度可以作为筛选大白经产发情母猪和返情母猪的测定部位（准确率分别为88%和81%）。因此，Y3TB01体温筛查智能摄像机和红外热像仪C3配套技术能准确监测母猪体温和异常体温报警，明确不同胎次不同生理时期大白母猪体表部位温度和体温分布，准确鉴定出发情母猪和返情母猪。研究为非接触式测温鉴定母猪发情技术提供了科学依据，对规模化猪场母猪的饲养管理与疫情防控具有重要意义。

Identification of the typical physiological state of sows by non-contact temperature measurement

Abstract: Body temperature is a key physiological indicator to monitor the disease and estrus in pigs. The estrus status of sows can be identified to assess the health status by the rectal temperature with a thermometer in the large-scale pig farms at present. However, traditional manual identification and temperature measurement cannot fully meet the high demand of large-scale production, particularly with the increase in labor costs under the harsh epidemic prevention and control level. Fortunately, infrared thermography can be expected to monitor the body temperature of sows for the identification of sows in estrus. It is obviously different in the body temperature of sows in estrus and interestrus. Specifically, the body temperature in estrus is significantly higher than that in interestrus. In this study, the non-contact temperature measurement was proposed to identify the typical physiological state of sows. A Y3TB01 smart camera was first used for body temperature screening. An infrared thermography camera C3 was then utilized to capture the infrared images of sows in real time. An image analysis software was finally selected to obtain the body surface temperature after the non-contact temperature measurement. A total of 720 primiparous and farrowing sows were used as the study objects in this case. The body surface and rectal temperatures were measured during the nulliparous period, estrus, 1-8, and 9-16 d of gestation. The body surface temperature of the sows was measured using infrared thermography equipment, ranging from the eyes, ears, cochlea, udder, vulva, and rump. An electronic thermometer was utilized to measure the temperature of the rectal area. A correlation analysis was then performed between the body surface and rectal temperature, in order to screen the key sites of estrus and return to estrus temperature in the sows of different litters. The results showed that better performance was achieved in the Y3TB01 smart camera of body temperature screening (correlation coefficient of 0.973), compared with the rectal temperature measurement. The body surface temperature of sows was also accurately monitored in real time. The infrared thermal camera C3 was used to detect the abnormal body surface temperature of sows. The different parts of the sows were determined in estrus and returned to estrus by infrared equipment and thermometer. The accuracy rates of the vulva measurement site were 88% and 81% in estrus and returning to estrus, respectively, whereas, those were only 77% and 72% in the rump measurement site, respectively. Overall, the combination of the Y3TB01 smart camera of body temperature screening and infrared thermal camera C3 accurately and rapidly monitored the body temperature of sows in real time. The Internet of Things was then integrated to transmit the data into the cell phone or computer terminal. The warning alarm of the abnormal body temperature was realized to monitor the health condition of pigs. In addition, the temperature distribution was obtained at the body surface sites of sows in different litters and periods. The finding can provide a scientific basis for the non-contact identification of sows in estrus and return to estrus, particularly for the intelligent feeding management and epidemic prevention and control of sows in large-scale pig farms.