Human Pose Detection Tool

Automatically detect human pose from uploaded images, extract keypoint information, and generate a visual pose diagram.

点击上传图片

Examples

Features

• Smart detection with an advanced deep learning model
• Visual skeleton and keypoint rendering for intuitive pose display
• JSON export for later analysis
• Pose image download
• High-precision detection for 17 major body keypoints

Supported keypoints

• Head: nose, left eye, right eye, left ear, right ear
• Upper body: left shoulder, right shoulder, left elbow, right elbow, left wrist, right wrist
• Lower body: left hip, right hip, left knee, right knee, left ankle, right ankle

How to use

1. Upload an image by clicking the upload area or dragging a file into it
2. Click Detect Pose and the system will analyze the image
3. View the generated pose image and detection information on the right
4. Download the pose image or export keypoint data as JSON

Use cases

• Sports analysis
• Fitness form checking
• Dance teaching
• Medical rehabilitation posture assessment
• Game development and character animation data
• Ergonomics and biomechanics research

Notes

• Use clear, well-lit images for best results
• The person should be as complete as possible and not heavily occluded
• Supports common image formats such as JPG, PNG, and WebP
• The first use loads a model from storage.googleapis.com, about 12.2MB, and may take a few seconds
• All processing runs locally in the browser and images are not uploaded to the server

Data format

{
  "timestamp": "2024-01-01T12:00:00.000Z",
  "score": 0.85,
  "keypoints": [
    {
      "part": "nose",
      "position": { "x": 125.5, "y": 80.2 },
      "score": 0.92
    }
  ]
}

• timestamp
• score
• keypoints