返回
Keypoint-Based Forest Musk Deer Behavioral Recognition Method
DOI:10.3390/ani16111594.png)
摘要
En 中文
传统林麝行为监测依赖人工观察或视频回放,存在效率低、主观性强及难以实时预警等问题,从而限制了繁育和保育工作的开展。为解决这些问题,本研究提出了一种基于改进YOLOv8-Pose的行为识别方法。通过构建涵盖四种行为的关鍵点数据集并引入优化模块,提升了目标检测和姿态估计的精度,性能优于多个现有模型。同时开发了可视化界面,为林麝的人工繁育和野外保育提供了高效、低成本的自动化分析工具,从而为濒危物种的智能化保护做出贡献。
Keyword:
musk deer
keypoint detection
improved YOLOv8-Pose
behavioral recognition
real-time monitoring
期刊
IF:
2.7
论文数:
6.3K
被引数:
5.1W
机构
引用论文
Automatic measurement method for sheep body dimensions based on posture compensation基于姿态补偿的羊体尺寸自动测量方法
Automated Measurement of Cattle Dimensions Using Improved Keypoint Detection Combined with Unilateral Depth Imaging基于改进的关键点检测与单侧深度成像相结合的牛体尺寸自动测量
Animals
IF0
Computation and analysis of phenotypic parameters of Scylla paramamosain based on YOLOv11-DYPF keypoint detection基于YOLOv11-DYPF关键点检测的斑节对虾表型参数计算与分析
A detection method for synchronous recognition of string tomatoes and picking points based on keypoint detection基于关键点检测的串收番茄与采摘点同步识别检测方法
An algorithm for detecting cow lameness based on ensemble learning of keypoint motion features一种基于关键点运动特征集成学习检测奶牛跛行的算法

