Hand-Detection

Hand Detection System utilizing deep learning library from MediaPipe by Google

CHECK OUT Live Video Output Below !!

Input

• live frame

Output

• Hand detected
  • Palm detection
  • 21 point detection

Algorithm

• Palm Detection

To detect initial hand locations, we utlize a single-shot detector model optimized for mobile real-time uses Detecting hands is a decidedly complex task: we use model to work across a variety of hand sizes with a large scale span (~20x) relative to the image frame and be able to detect occluded and self-occluded hands.

• Hand Landmark Model

After the palm detection over the whole image our subsequent hand landmark model performs precise keypoint localization of 21 3D hand-knuckle coordinates inside the detected hand regions via regression, that is direct coordinate prediction. The model learns a consistent internal hand pose representation and is robust even to partially visible hands and self-occlusions.

Result

Output Video

Hand.Detection.mp4