Modelling and Validation of Computer Vision Techniques to Assess Heart Rate, Eye Temperature, Ear-Base Temperature and Respiration Rate in Cattle

Abstract

Precision livestock farming has emerged with the aim of providing detailed information to detect and reduce problems related to animal management. This study aimed to develop and validate computer vision techniques to track required features of cattle face and to remotely assess eye temperature, ear-base temperature, respiration rate, and heart rate in cattle. Ten dairy cows were recorded during six handling procedures across two consecutive days using thermal infrared cameras and RGB (red, green, blue) video cameras. Simultaneously, core body temperature, respiration rate and heart rate were measured using more conventional ‘invasive’ methods to be compared with the data obtained with the proposed algorithms. The feature tracking algorithm, developed to improve image processing, showed an accuracy between 92% and 95% when tracking different areas of the face of cows. The results of this study also show correlation coefficients up to 0.99 between temperature measures obtained invasively and those obtained remotely, with the highest values achieved when the analysis was performed within individual cows. In the case of respiration rate, a positive correlation (r = 0.87) was found between visual observations and the analysis of non-radiometric infrared videos. Low to high correlation coefficients were found between the heart rates (0.09–0.99) obtained from attached monitors and from the proposed method. Furthermore, camera location and the area analysed appear to have a relevant impact on the performance of the proposed techniques. This study shows positive outcomes from the proposed computer vision techniques when measuring physiological parameters. Further research is needed to automate and improve these techniques to measure physiological changes in farm animals considering their individual characteristics.

Remotely Sensed Imagery for Early Detection of Respiratory Disease in Pigs: A Pilot Study

“…RGB videos were processes by the respective algorithm, which identifies changes in luminosity in the green colour channel within the ROI (eye area; Figure 1c) and applies a second-order Butterworth filter and a fast Fourier transformation (FFT) after the signal has been obtained. To improve the accuracy of this analysis, the ROI was tracked through a computer vision algorithm following the methodology suggested by Jorquera-Chavez et al [25]. This methodology first selects the most representative features within the ROI by using different pattern recognition techniques which then are tracked along with the sequential frames of the video.…”

“…Furthermore, for the analysis of respiration rate, non-radiometric infrared videos were processed using an algorithm that identifies the changes in pixel intensity values within the ROI (nose area; Figure 1d), which are related to the air exchange during exhalation and inhalation [25].…”

“…Collected images were processed using customised algorithms developed in Matlab ® R2018b (Mathworks Inc., Natick, MA, USA), which are described by Jorquera-Chavez et al [25]. In the case of TIR images, this algorithm firstly extracted the radiometric information of each image, by using FLIR ® Atlas SDK [25,26]. Secondly, it allowed the selection of the eye and ear-base as region(s) of interest (ROI(s)), from where the temperature were extracted (Figure 1a,b).…”

Is Continuous Heart Rate Monitoring of Livestock a Dream or Is It Realistic? A Review

“…Researchers are currently investigating non-contact computer-based techniques to monitor HR in livestock [119]. A low to high correlation coefficient (0.09-0.99) was found between HR obtained from the red, green and blue color channels (RGB) videos and from the gold standard method in cattle [120]. The study also showed that the accuracy of HR measurement is related to both the position of the cameras and the area analyzed on the images.…”

“…To date, many signal processing techniques have been proposed to remove MAs in PPG signals. Adaptive noise cancellation (ANC) is a popular approach to remove MAs where reference signals can be constructed from acceleration data or another PPG signal [39,[120][121][122][123][124][125][126][127][128][129][130][131][132][133]. However, the drawback is that the performance of ANC is sensitive to the reference signal, and it is difficult to reconstruct qualified reference signals during exercising.…”