Quality-Gated Episode-Level RGB-D Body Condition Scoring in Beef Cattle.
- 1College of Information Science and Engineering, Shanxi Agricultural University, Jinzhong 030801, China.
- 2Shanxi Province Science and Technology Information Research Center, Taiyuan 030027, China.
Abstract
In practical beef cattle monitoring scenarios, automated body condition scoring (BCS) based on single-frame visual information can be affected by posture variation, occlusion, and image quality changes. This study proposed an episode-level BCS assessment framework using synchronized RGB-D videos, integrating multi-frame visual information, attention-based multiple instance learning (Attention-MIL), and a dorsal geometry-based quality control mechanism to improve reliability under complex acquisition conditions. The publicly available Raramuri Criollo dataset was reorganized into a dataset containing 3815 RGB-D frame pairs, 52 animal identity groups, and 151 episodes, and evaluated using animal-wise three-fold cross-validation. Predictions were generated only when an episode contained at least three quality-filtered frames and reliable angle estimation. With an average coverage of 0.7345 after quality gating, the final model achieved 61.66% accuracy, a Macro-F1 score of 0.5680, and a quadratic weighted kappa (QWK) of 0.6339 for five-class BCS classification. Additionally, 89.71% of predictions were within one BCS category of the reference scores. Three-class management classification achieved a Macro-F1 score of 0.8059. Ablation experiments showed complementary contributions from RGB and depth information, and the combined angle and quality-statistic features improved several metrics in the original ablation. Additional matched experiments did not establish an independent angle benefit but supported multi-frame aggregation and within-Episode prediction consistency. The proposed RGB-D multi-frame fusion approach demonstrates potential for automated beef cattle body condition monitoring under controlled acquisition conditions.
Related
- CLSP-REQA: a real-time quality-aware closed-loop seizure prediction framework with Mamba-BiLSTM and confidence-gated intervention
- Peroxisome quality control is gated by a dynamic ATF6α-ABCD3 tether
- Evaluation of the Obstetric Quality of Recovery-10 scoring tool up to 30 days postpartum: a secondary cross-sectional psychometric analysis (SOCRATES cohort)
- Effects of castration-associated hormonal changes on rumen microbiota, metabolism, and meat quality in Holstein cattle.
- Meta-analysis with batch-effect adjustment reveals breed-specific ruminal microbiome variation of beef cattle and its implications for meat quality and productivity.
BibTeX
@article{hou2026qualitygated,
title = {{Quality-Gated Episode-Level RGB-D Body Condition Scoring in Beef Cattle.}},
author = {Zhen Hou and Weinan Cao and Wei He and Yanwen Li and Jiarui Li and Di Wu and Ruiping Wang and Hua Yang},
journal = {Animals : an open access journal from MDPI},
year = {2026},
month = sep,
doi = {10.3390/ani16193040},
url = {https://europepmc.org/article/MED/42883009},
note = {CC BY},
}
RIS
TY - JOUR
TI - Quality-Gated Episode-Level RGB-D Body Condition Scoring in Beef Cattle.
AU - Zhen Hou
AU - Weinan Cao
AU - Wei He
AU - Yanwen Li
AU - Jiarui Li
AU - Di Wu
AU - Ruiping Wang
AU - Hua Yang
PY - 2026
DA - 2026/09/27
JO - Animals : an open access journal from MDPI
DO - 10.3390/ani16193040
UR - https://europepmc.org/article/MED/42883009
AB - In practical beef cattle monitoring scenarios, automated body condition scoring (BCS) based on single-frame visual information can be affected by posture variation, occlusion, and image quality changes. This study proposed an episode-level BCS assessment framework using synchronized RGB-D videos, integrating multi-frame visual information, attention-based multiple instance learning (Attention-MIL), and a dorsal geometry-based quality control mechanism to improve reliability under complex acquisition conditions. The publicly available Raramuri Criollo dataset was reorganized into a dataset containing 3815 RGB-D frame pairs, 52 animal identity groups, and 151 episodes, and evaluated using animal-wise three-fold cross-validation. Predictions were generated only when an episode contained at least three quality-filtered frames and reliable angle estimation. With an average coverage of 0.7345 after quality gating, the final model achieved 61.66% accuracy, a Macro-F1 score of 0.5680, and a quadratic weighted kappa (QWK) of 0.6339 for five-class BCS classification. Additionally, 89.71% of predictions were within one BCS category of the reference scores. Three-class management classification achieved a Macro-F1 score of 0.8059. Ablation experiments showed complementary contributions from RGB and depth information, and the combined angle and quality-statistic features improved several metrics in the original ablation. Additional matched experiments did not establish an independent angle benefit but supported multi-frame aggregation and within-Episode prediction consistency. The proposed RGB-D multi-frame fusion approach demonstrates potential for automated beef cattle body condition monitoring under controlled acquisition conditions.
SN - 2076-2615
ER -