FemaleSaanenGoat
收藏资源简介:
FemaleSaanenGoat是由西北农林科技大学等机构联合创建的首个萨能奶山羊三维数据集,包含55只6-18月龄雌性山羊的八视角同步RGBD视频数据。数据集通过多视角动态融合算法处理,生成约3,200个高精度三维扫描模型(训练集2,700个/测试集500个),每个模型标注32个解剖学关键点。数据采集采用轻量化农场环境部署方案,捕捉山羊站立、行走等自然姿态的30秒30FPS视频序列。该数据集为构建物种专用参数化模型SaanenGoat提供基础,支持精准畜牧业中的三维重建、体型测量等应用,解决了现有动物模型在奶山羊形态表征上的精度缺陷问题。
The FemaleSaanenGoat dataset, jointly developed by Northwest A&F University and other institutions, is the first 3D dataset for Saanen dairy goats. It contains eight-view synchronized RGBD video data of 55 female goats aged 6 to 18 months. Processed using multi-view dynamic fusion algorithms, the dataset generates approximately 3,200 high-precision 3D scanning models, with 2,700 samples in the training set and 500 in the test set. Each model is annotated with 32 anatomical key points. The data collection adopts a lightweight farm environment deployment plan, capturing 30-second 30FPS video sequences of natural goat postures including standing and walking. This dataset lays a foundation for constructing the species-specific parametric model SaanenGoat, supports applications such as 3D reconstruction and body size measurement in precision livestock farming, and resolves the accuracy deficiencies of existing animal models in characterizing the morphology of Saanen dairy goats.
- 1Monocular Mesh Recovery and Body Measurement of Female Saanen Goats西北农林科技大学·信息工程学院; 南方科技大学·电子与电气工程系; 清华大学·北京国家信息科学与技术研究中心; 清华大学·自动化系; 农业农村部农业物联网重点实验室; 陕西省农业信息感知与智能服务重点实验室; 陕西省农业信息智能感知与分析工程研究中心 · 2026年



