The demand for efficient and precise body measurement systems has surged across diverse sectors. In these industries, accurate anthropometric data is crucial for applications ranging from personalized health monitoring and ergonomic assessment to custom apparel fitting and sports performance optimization. Traditional body measurement methods, while effective, present notable challenges: they are often laborintensive, time-consuming, and susceptible to human error, in addition to requiring specialized tools and skilled personnel for reliable results [12]. Existing automated systems often rely on complex 3D imaging setups or require extensive manual calibration, limiting their accessibility and practicality for widespread use [7]. As a result, there is a growing need for solutions that can deliver fast, accurate, and consistent measurements with minimal human intervention. Technologies that fully automate body measurements using 2D images minimize manual effort while maintaining or even enhancing measurement precision. Our method leverages 2D pose estimation to address these limitations, extracting anthropometric data efficiently without 3D setups or intensive computation. The novelty of our idea is its use of elliptical cross-sections for circumference calculations and the integration of MediaPipe for low-resource pose estimation This approach bypasses costly equipment and calibration, applying flexible 2D landmarks to automate body measurements across various devices, presenting a practical alternative for real-time, high-accuracy applications.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Image-Based Anthropometric Measurement System Using Pose Estimation

  • Vansh Pundir,
  • Ilesha Ojha,
  • G. Maragatham

摘要

The demand for efficient and precise body measurement systems has surged across diverse sectors. In these industries, accurate anthropometric data is crucial for applications ranging from personalized health monitoring and ergonomic assessment to custom apparel fitting and sports performance optimization. Traditional body measurement methods, while effective, present notable challenges: they are often laborintensive, time-consuming, and susceptible to human error, in addition to requiring specialized tools and skilled personnel for reliable results [12]. Existing automated systems often rely on complex 3D imaging setups or require extensive manual calibration, limiting their accessibility and practicality for widespread use [7]. As a result, there is a growing need for solutions that can deliver fast, accurate, and consistent measurements with minimal human intervention. Technologies that fully automate body measurements using 2D images minimize manual effort while maintaining or even enhancing measurement precision. Our method leverages 2D pose estimation to address these limitations, extracting anthropometric data efficiently without 3D setups or intensive computation. The novelty of our idea is its use of elliptical cross-sections for circumference calculations and the integration of MediaPipe for low-resource pose estimation This approach bypasses costly equipment and calibration, applying flexible 2D landmarks to automate body measurements across various devices, presenting a practical alternative for real-time, high-accuracy applications.