An Accurate Positioning Method of Image Target Based on Sub-Pixel
Jian Lin, Gang Fu, Xue Lu Yao · Applied Mechanics and Materials · 2014
The conventional image of the mark is mainly targeting the video image within the target pixel level positioning, the positioning accuracy is poor, and is easily affected by noise. To this end, the introduction of a Canny edge detection and Zernike moments method of combining precise positioning mark targets for sub-pixel edge detection, and for the oval shape characteristics mark the target, using the least squares fitting to mark the target oval central location positioning . Experimental results show that the higher positioning accuracy of this method, anti-interference ability, with high robustness, and can meet the requirements of real-time video image processing.