An adaptive video stream sampling algorithm for improving the detection accuracy of reference anatomical landmarks in cattle for their body condition assessment

Body Condition Score (BCS) in dairy cattle is used to monitor animal energy status and to identify the risks of metabolic and reproductive disorders. In industrial video recording, the accuracy of automated BCS estimation is limited by the displacement of reference anatomical landmarks. This displacement arises due to changes in animal posture, walking speed, uneven lighting, glare, and shadows. The study aimed to develop an adaptive video-frame selection algorithm to improve the detection accuracy of reference keypoints and contour landmarks in automated body condition assessment of dairy cattle. The research material consisted of video sequences of 983 dairy cows. For analysis, a system of reference points and contour landmarks of the pelvic region was used. Image processing was performed using a multitask neural network model for localizing key points and contours, followed by BCS prediction on a 1-5 scale. The developed algorithm included adaptive frame sampling based on animal walking speed, visibility assessment of anatomical zones, and component-wise selection of the most informative frames to form a consistent set of features. Application of the algorithm resulted in a reduction of the mean absolute error in BCS estimation from 0.34 to 0.22 score points, an increase in the proportion of predictions within ±0.5 points from 86.2% to 93.7%, and an improvement in the weighted Cohen’s κ coefficient from 0.74 to 0.86. The normalized localization error of reference keypoints decreased from 0.071 to 0.048. The findings confirm that adaptive frame preprocessing enhances the accuracy and robustness of automated body condition assessment under industrial video recording conditions.

1. Sadikov R.Z., Morozova N.I., Sadikov R.R., Morozov I.A. et al. The system of automatic determination of fatness of cows as a tool for maintaining optimal physiological health and productivity. Bulletin of the Ryazan State Agrotechnological University named after P.A. Kostychev. 2024;16(1):54-61. (In Russ.) https://doi.org/10.36508/RSATU.2024.83.84.008

2. Turlyun V.I. Body condition score of the Holstein dairy cattle. Sbornik nauchnykh trudov Severo-Kavkazskogo nauchno-issledovatelskogo instituta zhivotnovodstva. 2014;1(3):167-170. (In Russ.)

3. Loker S., Bastin C., Miglior F., Sewalem A. et al. Genetic and environmental relationships between body condition score and milk production traits in Canadian Holsteins. Journal of Dairy Science. 2012;95(1):410-419. https://doi.org/10.3168/jds.2011-4497

4. Gentil N.R., Santos A.D.F., Pereira M.A., Lima M.S. et al. Validation of body condition score system in Holstein × Zebu cows via ultrasound method during the transition period and early lactation. Semina: Ciências Agrárias. 2017;38(6):3797-3806. https://doi.org/10.5433/1679-0359.2017v38n6p3797

5. Nagy S.Á., Kilim O., Csabai I., Gábor G. et al. Impact Evaluation of Score Classes and Annotation Regions in Deep Learning-Based Dairy Cow Body Condition Prediction. Animals. 2023;13(2):194. https://doi.org/10.3390/ani13020194

6. Bewley J.M., Schutz M.M. An interdisciplinary review of body condition scoring for dairy cattle. The Professional Animal Scientist. 2008;24(6):507-529. https://doi.org/10.15232/S1080-7446(15)30901-3

7. Morrow D.A. Fat cow syndrome. Journal of Dairy Science. 1976;59(9):1625-1629. https://doi.org/10.3168/jds.S0022-0302(76)84415-3

8. Roche J.R., Kay J.K., Friggens N.C., Loor J.J. et al. Assessing and managing body condition score for the prevention of metabolic disease in dairy cows. Veterinary Clinics of North America: Food Animal Practice. 2013;29(2):323-336. https://doi.org/10.1016/j.cvfa.2013.03.003

9. Bezborodov P.N. Condition parameter “Body Condition Score (BCS)” of the dairy cows with displaced abomasum. Vestnik NSAU (Novosibirsk State Agrarian University. 2017;(4):124-141. (In Russ.)

10. Roche J.R., Friggens N.C., Kay J.K., Fisher M.W. et al. Invited review: Body condition score and its association with dairy cow productivity, health, and welfare. Journal of Dairy Science. 2009;92(12):5769-5801. https://doi.org/10.3168/jds.2009-2431

11. Nenashev I.V., Marin E.M. The prevalence of orthopedic diseases of the distal extremities in cattle depending on housing conditions. Bulletin Samara State Agricultural Academy. 2004;9(4):103-108. (In Russ.) https://doi.org/10.55170/1997-3225-2024-9-4-103-108

12. Sun Y., Huo P., Wang Y., Cui Z. et al. Automatic monitoring system for individual dairy cows based on a deep learning framework that provides identification via body parts and estimation of body condition score. Journal of Dairy Science. 2019;102(11):10140-10151. https://doi.org/10.3168/jds.2018-16164

13. Summerfield G.I., De Freitas A., van Marle-Koster E., Myburgh H.C. Automated Cow Body Condition Scoring Using Multiple 3D Cameras and Convolutional Neural Networks. Sensors. 2023;23(22):9051. https://doi.org/10.3390/s23229051

14. Çevik K.K., Boğa M. Body Condition Score (BCS) Segmentation and Classification in Dairy Cows Using R-CNN Deep Learning Architecture. European Journal of Science and Technology. 2019;17:1248-1255. https://doi.org/10.31590/ejosat.658365

15. Zin T.T., Seint P.T., Tin P., Horii Y. et al. Body Condition Score Estimation Based on Regression Analysis Using a 3D Camera. Sensors. 2020;20(13):3705. https://doi.org/10.3390/s20133705

16. Mishchenko I.I., Misnik A.E., Alexandrov A.V. Application of computer vision and image preprocessing technologies in decision support systems. Vestnik Samarskogo gosudarstvennogo tekhnicheskogo universiteta. Seriya “Tekhnicheskie nauki”. 2024;32(4):6-26. (In Russ.) https://doi.org/10.14498/tech.2024.4.1

17. Olt L.S., Rice M.S., St. Clair M., Million C. et al. III. Automating Region of Interest (ROI) Selection for Spectral Analyses of Mastcam-Z Observations. 56th Lunar and Planetary Science Conference. 2025:2154. URL: (accessed: Marhttps://www.hou.usra.edu/meetings/lpsc2025/pdf/2154.pdf ch 13, 2026).

18. PyTorch Contributors. AdamW – PyTorch Documentation. URL: (accessed: March https://docs.pytorch.org/docs/stable/generated/torch.optim.AdamW.html 13, 2026).