Lameness has a considerable influence on the welfare and health of dairy cows. Many attempts have been made to develop automatic lameness detection systems using computer vision technology. However, these detection methods are easily affected by the characteristics of individual cows, resulting in inaccurate detection of lameness. Therefore, this study explores an individualized lameness detection method for dairy cattle based on the supporting phase using computer vision. This approach is applied to eliminate the influence of the characteristics of individual cows and to detect lame cows and lame hooves. In this paper, the correlation coefficient between lameness and the supporting phase is calculated, a lameness detection algorithm based on the supporting phase is proposed, and the accuracy of the algorithm is verified. Additionally, the reliability of this method using computer vision technology is verified based on deep learning. One hundred naturally walking cows are selected from video data for analysis. The results show that the correlation between lameness and the supporting phase was 0.864; 96% of cows were correctly classified, and 93% of lame hooves were correctly detected using the supporting phase-based lameness detection algorithm. The mean average precision is 87.0%, and the number of frames per second is 83.3 when the Receptive Field Block Net Single Shot Detector deep learning network was used to detect the locations of cow hooves in the video. The results show that the supporting phase-based lameness detection method proposed in this paper can be used for the detection and classification of cow lameness and the detection of lame hooves with high accuracy. This approach eliminates the influence of individual cow characteristics and could be integrated into an automatic detection system and widely applied for the detection of cow lameness.

The computer vision technique has been rapidly adopted in cow lameness detection research due to its noncontact characteristic and moderate price. This paper attempted to summarize the research progress of computer vision in the detection of lameness. Computer vision lameness detection systems are not popular on farms, and the accuracy and applicability still need to be improved. This paper discusses the problems and development prospects of this technique from three aspects: detection methods, verification methods and application implementation. The paper aims to provide the reader with a summary of the literature and the latest advances in the field of computer vision detection of lameness in dairy cows.

Weight is an important indicator of the growth and development of dairy cows. The traditional static weighing methods require considerable human and financial resources, and the existing dynamic weighing algorithms do not consider the influence of the cow motion state on the weight curve. In this paper, a dynamic weighing algorithm for cows based on a support vector machine (SVM) and empirical wavelet transform (EWT) is proposed for classification and analysis. First, the dynamic weight curve is obtained by using a weighing device placed along a cow travel corridor. Next, the data are preprocessed through valid signal acquisition, feature extraction, and normalization, and the results are divided into three active degrees during motion for low, medium, and high grade using the SVM algorithm. Finally, a mean filtering algorithm, the EWT algorithm, and a combined periodic continuation-EWT algorithm are used to obtain the dynamic weight values. Weight data were collected for 910 cows, and the experimental results displayed a classification accuracy of 98.6928%. The three algorithms were used to calculate the dynamic weight values for comparison with real values, and the average error rates were 0.1838%, 0.6724%, and 0.9462%. This method can be widely used at farms and expand the current knowledgebase regarding the dynamic weighing of cows.

Aiming at the problem of low non-invasive intelligent identification of pigs in healthy breeding, a pig individual identification method based on the improved YOLOv4 convolutional neural network is proposed. By preprocessing the images of 30 live pigs in the collected video, a method based on the whole live pig as the feature extraction sample is proposed, instead of using the pig face as the feature extraction sample, to make a data set to obtain the effective features of the whole body. Considering the living habits of live pigs and the living environment of live pigs, the improved YOLOv4 model is proposed, which adapted certain data enhancement and attention mechanisms, and conducts multi-scale training and detection on the data set to achieve accurate detection and identification of individual live pigs; obtains the best parameters according to the training log, and carries out verification and comparison experiments for individual pig identification. The result shows that when the threshold is 0.5, the average accuracy (mAP) on the test set reaches 98.12%. The recall rate reached 95%, the F1-score was 96%, the average recognition time Mean FPS was 34.3ms, and the average crossover ratio (IoU) was 83.91%. The improved model comparison between YOLOv3 and Fast R-CNN has improved the recognition accuracy and speed of individual pigs; at the same time, compared with the traditional pig face data set, the pig body data set has significantly improved the recognition accuracy. Based on the whole live pig as a feature extraction sample combined with the improved YOLOv4 model, the live pig can be recognized quickly and effectively in various postures and complex backgrounds. As a multi-angle and multi-target recognition method, it can quickly and accurately identify individual pigs in a large-scale group healthy breeding environment as well as realization of individualized healthy breeding and efficient management for pig.

Lameness significantly compromises dairy cattle welfare and productivity. Early detection enables prompt intervention, enhancing both animal health and farm efficiency. Current computer vision approaches often rely on isolated lameness feature quantification, disregarding critical interdependencies among gait parameters. This limitation is exacerbated by the distinct kinematic patterns exhibited across lameness severity grades, ultimately reducing detection accuracy. This study presents an integrated computer vision and deep-learning framework for dairy cattle lameness detection and severity classification. The proposed system comprises (1) a Cow Lameness Feature Map (CLFM) model extracting holistic gait kinematics (hoof trajectories and dorsal contour) from walking sequences, and (2) a DenseNet-Integrated Convolutional Attention Module (DCAM) that mitigates inter-individual variability through multi-feature fusion. Experimental validation utilized 3150 annotated lameness feature maps derived from 175 Holsteins under natural walking conditions, demonstrating robust classification performance. The classification accuracy of the method for varying degrees of lameness was 92.80%, the sensitivity was 89.21%, and the specificity was 94.60%. The detection of healthy and lameness dairy cows’ accuracy was 99.05%, the sensitivity was 100%, and the specificity was 98.57%. The experimental results demonstrate the advantage of implementing lameness severity-adaptive feature weighting through hierarchical network architecture.