In order to achieve accurate and efficient recognition of calf behavior in complex scenes such as cow overlapping, occlusion, and different light and occlusion levels, this experiment adopts the method of improving the YOLO v8 model to recognize calf behavior. A calf daily behavior dataset containing 2918 images is selected as the test benchmark through video frame extraction; a P2 small-target detection layer is introduced to improve the resolution of the input scene, which significantly improves the model recognition accuracy, and reduces the computational complexity and storage requirements of the model through the Lamp pruning method. Comparisons are made with the SSD, YOLOv5n, YOLOv8n, YOLOv8-C2f-faster-EMA, YOLO v11n, YOLO v12n, and YOLO v8-P2 advanced models. The results show that the number of parameters, floating point operations (FLOPs), model size, and mean average precision (mAP) of the model after introducing the P2 small-target detection layer and pruning with the Lamp strategy are 0.949 M, 4.0 G, 2.3 Mb, and 90.9%, respectively. The significant improvement in each index effectively reduces the model size and improves the accuracy of the network. The detection results in complex environments with different light and shading levels show that the mAP in daytime (exposure) and nighttime environments is 85.1% and 84.8%, respectively, and the average mAP in the three kinds of shading cases (light, medium, and heavy) is 87.3%, representing a lightweight, high-precision, real-time, and robust model. The results of this study provide a reference for the real-time monitoring of calf behaviors all day long in complex environments.

The daily behaviors of dairy cows, including standing, drinking, eating, and lying down, are closely associated with their physical health. Efficient and accurate recognition of dairy cow behaviors is crucial for timely monitoring of their health status and enhancing the economic efficiency of farms. To address the challenges posed by complex scenarios and significant variations in target scales in dairy cow behavior recognition within group farming environments, this study proposes an enhanced recognition method based on YOLOv5. Four Shuffle Attention (SA) modules are integrated into the upsampling and downsampling processes of the YOLOv5 model’s neck network to enhance deep feature extraction of small-scale cow targets and focus on feature information, while maintaining network complexity and real-time performance. The C3 module of the model was enhanced by incorporating Deformable convolution (DCNv3), which improves the accuracy of cow behavior characteristic identification. Finally, the original detection head was replaced with a Dynamic Detection Head (DyHead) to improve the efficiency and accuracy of cow behavior detection across different scales in complex environments. An experimental dataset comprising complex backgrounds, multiple behavior categories, and multi-scale targets was constructed for comprehensive validation. The experimental results demonstrate that the improved YOLOv5 model achieved a mean Average Precision (mAP) of 97.7%, representing a 3.7% improvement over the original YOLOv5 model. Moreover, it outperformed comparison models, including YOLOv4, YOLOv3, and Faster R-CNN, in complex background scenarios, multi-scale behavior detection, and behavior type discrimination. Ablation experiments further validate the effectiveness of the SA, DCNv3, and DyHead modules. The research findings offer a valuable reference for real-time monitoring of cow behavior in complex environments throughout the day.