Recognition of Printed Character on a plate Based on Feature Weighted Template Matching
Zhang Jing-y · Journal of Shenyang Ligong University · 2015
According to recognition problem of printed characters on a plate,an uneven brightness and uneven printed characters recognition algorithm is presented. The Top-hat transformation and filtering was used to remove image noise in the image preprocessing; the segmentation algorithm based on OTSU method was used to achieve the binarization of image; the width and height of the image and the projection method were used to split the characters; weight template matching algorithm based on feature was used to recognize the character on the plate. The accurate rate of printed characters on the plate has reached 98. 00%through doing experiments of the plate printing image captured by the camera. The effectiveness of the algorithm is proved.