Research of Hot Stamping Image Recognition Algorithm Based on Projection Feature

To meet the requirements of real-time on-line detection, an algorithm for hot stamping image recognition was proposed based on projection with adaptive threshold. Firstly CCD collects digital image, then convert RGB images into HSV space, and extract V channel. The initial threshold was set according to the error on the first location of matching region. In the subsequent template matching process, the threshold would be update; the smallest D-value point is the best match. The experiment results show that the improved algorithms can identify hot stamping area and greatly increase the matching velocity, having certain guiding significance to hot stamping quality detection based on machine vision.