Intelligent monitoring method of cow ruminant behavior based on video analysis technology

To overcome the limitations of traditional dairy cow's rumination detection methods, a video-based analysis on the intelligent monitoring method of cow ruminant behavior was proposed in this study. The Mean Shift algorithm was used to track the jaw motion of dairy cows accurately. The centroid trajectory curve of the cow mouth motion was subsequently extracted from the video. In this way, the monitoring of the ruminant behavior of dairy cows was realized. To verify the accuracy of the method, six videos, a total of 99'00", 24 000 frames were selected. The test results demonstrated that the success rate of this method was 92.03%, despite the interference of behaviors, such as raising or turning of the cow’s head. The results demonstrate that this method, which monitors the ruminant behavior of dairy cows, is effective and feasible.