Binocular Stereo Vision Ranging Algorithm

This paper proposes an algorithm based on the principle of binocular stereo-vision to measure the distance from the camera to the finger. The algorithm which is used in wearable computer system for human-computer interaction, takes a fingertip as a feature matching point to locate the finger space coordinates based on the method of triangulation range. Proposes a finger detection algorithm based on fitting an ellipse to the finger contour and according to the fitting elliptical direction and the main spindle length to calculate the pixel coordinates of the fingertip in the image and use HALCON on a finger detection algorithm research, simulation and parameter optimization. Uses the epipolar constraint and feature point neighborhood statistics character as the matching rules to calculate the parallax which used to calculate three-dimensional space coordinates of the fingertip. Constructed the experimental platform and completed ranging system software design, which could locate the space finger real-time. Experimental results show that the ellipse fitting fingertips detection algorithm can effectively extracted a finger from the background and the distance gets form the ranging algorithm basically consistent with the measured distance.