Pig Body Temperature Detection Method Based On Infrared Image

Body temperature measurement is a very important task in the sow breeding process.The traditional sow body temperature measurement method uses a mercury column to measure the rectal temperature of the sow,which is an invasive measurement method.The temperature measurement process causes stress in the sow,which is not conducive to the healthy growth and breeding of the sow.Infrared temperature measurement technology is gradually emerging in modern.Infrared temperature measurement gun is an infrared temperature measurement device based on point analysis.It measures the point temperature of the target body,and the measurement point selection is random,which is easy to cause measurement inaccuracy.The infrared image is a temperature measurement technology based on temperature field data analysis,which can invert the temperature distribution of the body surface of the pig.The scholars mainly focus on the use of infrared images to obtain the distribution of the temperature field data of the animal thermal infrared image and the highest temperature value and average temperature value,which mostly used for health assessment,disease detection,breeding and breeding,etc.In this paper,the infrared camera is used to identify the temperature of the body surface of the breeding sow through the average value of the infrared image temperature in the ear root region on the basis of identifying the root region of the sow’s ear.Compared with the infrared temperature gun,the method can overcome the randomness of the measurement point selection,and analyze the regional temperature,and the body temperature data acquisition is more comprehensive and accurate.Infrared image of the sow’s head(including the ear root area)was obtained using a Fortric-225 infrared camera at an ambient temperature of 28-33 °C.Using the gray value and temperature value corresponding to 180 infrared images,a T-G(Temperature-Gray)model based on the gray value of the target image of the infrared image and the corresponding temperature value was established by linear least squares method,which achieved temperature inversion of each pixel of the target pig.For the different growth stages and different breeds of sows,the R2 of the established model is >0.9.And randomly selected 60 infrared images of sows,based on the temperature of each pixel point derived from the software attached to the thermal infrared camera.The average relative error of the model inversion of the body temperature was only 0.076977%.Based on the T-G model,this paper designs a kind of sow’s ear root recognition and body surface temperature detection algorithm for different sows population scenarios.According to the gray scale range calibration of the sows body and the image exclusive OR operation,the root region of each sows is automatically cut,and the gray data of the ear root region of each sows is extracted.The temperature of the ear root region was calculated using the T-G model,and the mean value was calculated,which is the body surface temperature of the individual sows.Twenty four-month-old Rongchang sows were measured for their rectal temperature by a mercury thermometer,and the correlation between the body temperature measured by the algorithm and rectal temperature of the sows was analyzed It was found that there was a significant correlation between the body surface temperature and the rectal temperature at 0.01 level,and the correlation coefficient was 0.847825.Taking the temperature of each pixel point derived from the software attached to the thermal infrared camera as the reference standard,20 sets of sow’s infrared images and temperature measuring guns were used to measure the data.The sow’s body temperature measured by the algorithm and the sow’s body temperature measured by infrared thermometer were compared and analyzed.It was found that the average relative error of the body surface temperature detected by the algorithm is 0.132889%,and the average relative error of the infrared temperature measuring gun to measure the temperature of the ear root is 2.712676%.It can be seen that the algorithm for measuring the body temperature of the sow’s ear root region is more accurate than the infrared temperature gun based on the temperature measurement of the point temperature.For sows of different ages and different breeds,using this method,50 infrared images of 24-month-old Rongchang sows,50 infrared images of 24-month-old Landrace sows,50 infrared images of 4-month-old Rongchang sows,and 10 infrared images of 4-month-old Rongchang sows in multiple sow scenes were processed,the average relative error was 0.094479%,0.106143%,0.115799%,0.129756% respectively.According to the study of the influence of ambient temperature on the body surface temperature of sows,it was found that the ambient temperature had a significant effect on the body surface temperature of sows: Under normal circumstances,the higher the ambient temperature,the higher the body surface temperature of sows.At the same ambient temperature,the average body temperature of Rongchang sows varied from month to month.The body temperature of Rongchang sows at 4 months were higher than 24 months old Rongchang sows.Under the same environmental temperature,the average body temperature of different breeds of sows were different.The body temperature of 24 months old Rongchang sows were higher than 24 months old Landrace sowsOn the one hand,compared with the temperature measuring gun,the method of sows body surface temperature detection designed by this paper had higher precision and stronger stability.Furthermore,the algorithm could intelligently acquire the body surface temperature of the breeding sows without relying on the infrared camera temperature analysis software.This paper could provide a new way and technology for the accurate detection of body surface temperature values of sows in multiple scenes of sows using infrared camera.