Glossary

Pose Estimation

The identification of body landmarks and joint positions in images or video.

Pose estimation is a computer vision technique used to locate anatomical landmarks, such as the shoulders, hips, knees, and ankles, in an image or video. Connecting these landmarks creates a digital representation of body position, often called a skeleton.

In movement analysis, pose estimation can support joint tracking, angle visualization, and frame-by-frame observation without requiring physical markers to be attached to the body.

Factors That Affect Pose Estimation

  • Lighting, image resolution, and motion blur
  • Clothing and contrast with the background
  • Body parts being hidden or overlapping
  • Camera angle and distance
  • Fast or unusual movement positions

Pose-estimation output is an estimate rather than a direct measurement of internal joint structures. Landmark positions should be visually checked, especially when they are used to interpret technique or compare recordings.

Related Terms