Association of body condition with lameness in dairy cattle: a single-farm longitudinal study

Development of a real-time cattle lameness detection system using a single side-view camera

Recent advancements in machine learning and deep learning have revolutionized various computer vision applications, including object detection, tracking, and classification. This research investigates the application of deep learning for cattle lameness detection in dairy farming. Our study employs image processing techniques and deep learning methods for cattle detection, tracking, and lameness classification. We utilize two powerful object detection algorithms: Mask-RCNN from Detectron2 and the popular YOLOv8. Their performance is compared to identify the most effective approach for this application. Bounding boxes are drawn around detected cattle to assign unique local IDs, enabling individual tracking and isolation throughout the video sequence. Additionally, mask regions generated by the chosen detection algorithm provide valuable data for feature extraction, which is crucial for subsequent lameness classification. The extracted cattle mask region values serve as the basis for feature extraction, capturing relevant information indicative of lameness. These features, combined with the local IDs assigned during tracking, are used to compute a lameness score for each cattle. We explore the efficacy of various established machine learning algorithms, such as Support Vector Machines (SVM), AdaBoost and so on, in analyzing the extracted lameness features. Evaluation of the proposed system was conducted across three key domains: detection, tracking, and lameness classification. Notably, the detection module employing Detectron2 achieved an impressive accuracy of 98.98%. Similarly, the tracking module attained a high accuracy of 99.50%. In lameness classification, AdaBoost emerged as the most effective algorithm, yielding the highest overall average accuracy (77.9%). Other established machine learning algorithms, including Decision Trees (DT), Support Vector Machines (SVM), and Random Forests, also demonstrated promising performance (DT: 75.32%, SVM: 75.20%, Random Forest: 74.9%). The presented approach demonstrates the successful implementation for cattle lameness detection. The proposed system has the potential to revolutionize dairy farm management by enabling early lameness detection and facilitating effective monitoring of cattle health. Our findings contribute valuable insights into the application of advanced computer vision methods for livestock health management.

Diagnosis of lameness via data mining algorithm by using thermal camera and image processing method in Brown Swiss cows

Lameness is one of the culling factors such as mastitis, low milk yield, and infertility that cause economic losses in herd management as they threaten animal health and welfare. The purpose of this study was to evaluate the early detection of lameness in Brown Swiss cattle by using a data mining algorithm by both integrating lameness scores and some image parameters such as Lab (CIE L*, a*, b*), HSB (hue, saturation, brightness), RGB (red, green, blue) by processing thermal images with ImageJ program. In the study, the variables obtained as a result of processing the skin surface temperatures and thermal images taken at the fetlock joint of 33 Brown Swiss cattle were used as independent variables. Also, healthy cows (lameness scores 1 and 2) and unhealthy cows (lameness scores 3, 4, and 5) used in the diagnosis of lameness were used as a binary response variable. Classification and regression tree (CART) was used as a data mining algorithm in the diagnosis of lameness. As a result, the CART algorithm correctly classified 12 of the 13 heads unhealthy cows according to locomotion scores. According to locomotion scores by using CART analysis in this study, independent variables that are used to classify healthy and unhealthy (lame) animals were determined as maximum temperature (Tmax), green (mean), L (max), and age (P<0.05). The cut-off values of these independent variables were predicted as 32.40, 149.14, 97.11, and 5.50 for Tmax, green (mean), L (max), and age, respectively. Also, the sensitivity, specificity, and area under the ROC curve (AUC) of the CART algorithm for locomotion scoring were found as 92.31%, 95%, and 93.7% respectively. The area under ROC curve (AUC) was found to be significant in the diagnosis of lameness (P<0.01). Results showed that the use of CART classification algorithm together with thermal camera and image processing methods is a usefull tool in the detection of lameness in the herds. It is recommended that more comprehensive studies by increasing the number of animals in the future would be more beneficial.

Subclinical mastitis in dairy cows in south-Asian countries: a review of risk factors and etiology to prioritize control measures

Mastitis is a major production disease, causing significant economic losses for dairy farmers in South-Asian countries, as well as other parts of the world. Udder health control programs (UHCP) have been established in developed countries as an effective strategy for mastitis control but have not yet been introduced in South-Asian low-income countries like Bangladesh, India, Pakistan, and Sri Lanka. To launch UHCP successfully in dairy herds in South-Asia, it is important to know the current prevalence and risk factors for subclinical mastitis (SCM). Therefore, a narrative literature review was conducted with the aim to describe the dairy sector, the prevalence of SCM and its causal agents, risk factors for mastitis occurrence and the control measures suggested by different studies conducted in the selected countries. The literature revealed that India had the highest cattle population. Milking was mainly done by hand in all of the studied countries. Stall feeding was done in Bangladesh and Sri Lanka and limited access to grazing was also reported in some farms in India and Pakistan. There was substantial variation in the prevalence of SCM between studies in all 4 countries, ranging from about 20% to about 80%, but the average prevalence across all studies was high (50%). The most common causal agents for SCM were non-aureus staphylococci (NAS), Staphylococcus (S.) aureus, Streptococcus spp. and Escherichia (E.) coli. The management related risk factors reported for SCM were stall feeding of cows, a higher stock density, cracked floors, open drains, the presence of flies, poor drainage, peri-parturient diseases, infrequent dung removal and earth floors. The control measures suggested in these studies were to improve the hygiene and sanitation of cows, to improve the cleanliness of farms and milker's hands, to apply dry cow therapy, supplementing micronutrients and routine screening for SCM combined with taking intervention measures like isolation of cows or milking infected cows last, and proper treatment. Also, full hand milking, complete milking, machine milking, and providing feed and water immediately after milking have been recommended. Finally, we show that current literature often studies the same set of (non-manageable) risk factors, so more research is needed to obtain a comprehensive picture of the determinants of SCM. Randomized controlled trials are needed to truly quantify the effect of intervention under field conditions. Altogether, our work gives an overview of the udder health situation in South-Asia and provides the basis for the design of UHCP in this region.