基于融合图像与运动量的奶牛行为识别方法

为从海量监控视频中快速、准确识别影响奶牛繁殖与健康的行为,以小育成牛舍与泌乳牛舍中400头奶牛为研究对象,分析了奶牛在活动区与奶厅匝道的运动行为,提出了一种基于图像熵的奶牛目标对象识别方法,通过最小包围盒面积计算与目标对象轮廓图,实时捕获奶牛爬跨行为与蹄部、背部特征,融合被识别奶牛连续7 d的运动量,判断影响奶牛健康繁殖的异常行为。试验结果表明,利用本文方法对监控视频内奶牛目标对象、运动行为进行实时监测,有效监控识别奶牛发情、蹄病行为准确率超过80%,发情漏检率最低为3.28%,蹄病漏检率最低为5.32%,提高了规模化养殖管理效率。

Recognition Method of Cow Behavior Based on Combination of Image and Activities

Due to the application of internet of things (IoT) to large scale cow breeding, mass of multi scale data and multi divisional sensor data and video monitoring data of cow individuals were collected. Therefore, it is significant to dig out useful information about features of healthy reproduction behavior for development of scientific large scale breeding measures and improve economic benefits from cow breeding. For the rapid and accurate identification of cow reproduction and healthy behavior from mass surveillance video, totally 400 head of young cows and lactating cows were taken as the research object and cow behavior from the dairy activity area and milk hall ramp was analyzed. The method of object recognition based on image entropy was proposed, aiming at the identification of motional cow object behavior against a complex background. Calculation of a minimum bounding box and contour mapping was used for the real time capture of rutting span behavior and hoof or back characteristics. Then, by combining the continuous image characteristics with movement of cows for 7d, abnormal behavior of dairy cows from healthy reproduction can be quickly distinguished by the method, which improved the accuracy of the identification of dairy cows characteristics. Cow behavior recognition based on image analysis and activities was proposed to capture abnormal behavior that had harmful effects on healthy reproduction and improve the accuracy of cow behavior identification. The experimental results showed that through target detection, classification and recognition, the recognition rates of hoof disease and heat in the reproduction and health of dairy cows were greater than 80%, and the false negative rates of oestrus and hoof disease reached 3.28% and 5.32%, respectively. This method can enhance the real -time monitoring of cows, save time and improve the management efficiency of large scale farming.