Pose Estimation and Behavior Classification of Jinling White Duck Based on Improved HRNet

SIMPLE SUMMARY: The behavior and pose of ducks during the breeding process are directly related to their welfare and health status. The real-time and accurate monitoring of their different behaviors and poses is significant. In this study, we proposed a duck pose estimation method using computer vision techniques, achieving accurate detection of eight keypoints, including the head, beak, chest, left leg, right leg, left foot, right foot, and tail, under six behaviors: standing, drinking, preening, spreading wings, walking, and resting, and forming correct body expressions. Furthermore, the model’s generalization capability was tested with different lighting intensity levels. Ducks of various ages, breeds, and farming methods were used to validate the model’s comprehensive pose estimation capability. The model’s comprehensive detection capability was compared with mainstream pose estimation methods, proving its superiority. The results of this study are considered instrumental in smart poultry farming. ABSTRACT: In breeding ducks, obtaining the pose information is vital for perceiving their physiological health, ensuring welfare in breeding, and monitoring environmental comfort. This paper proposes a pose estimation method by combining HRNet and CBAM to achieve automatic and accurate detection of duck’s multi-poses. Through comparison, HRNet-32 is identified as the optimal option for duck pose estimation. Based on this, multiple CBAM modules are densely embedded into the HRNet-32 network to obtain the pose estimation model based on HRNet-32-CBAM, realizing accurate detection and correlation of eight keypoints across six different behaviors. Furthermore, the model’s generalization ability is tested under different illumination conditions, and the model’s comprehensive detection abilities are evaluated on Cherry Valley ducklings of 12 and 24 days of age. Moreover, this model is compared with mainstream pose estimation methods to reveal its advantages and disadvantages, and its real-time performance is tested using images of 256 × 256, 512 × 512, and 728 × 728 pixel sizes. The experimental results indicate that for the duck pose estimation dataset, the proposed method achieves an average precision (AP) of 0.943, which has a strong generalization ability and can achieve real-time estimation of the duck’s multi-poses under different ages, breeds, and farming modes. This study can provide a technical reference and a basis for the intelligent farming of poultry animals.