Application of Machine Vision in Defects Inspection and Character Recognition of Nameplate Surface

摘要

Aiming at a lot of weaknesses in manual inspection for defects inspection and characters recognition of nameplate, the paper proposed an application about inspecting and recognition of nameplate based on machine vision. This article firstly utilizes seed algorithm to process image, and remove incomplete nameplate image caused by shooting in order to retain nameplate's integrity. Secondly, it uses BLOB analysis algorithm to achieve precise detection for printing defect. Lastly, it adopts character recognize algorithm to divide region of character and count the ratio of white pixel to black pixel, accordingly attaining effective recognition of characters. Experiments show that the method of machine vision in defects inspection and character recognition of nameplate surface has faster inspection speed and higher precise, also commendably meets the requests of actual production.