Application of Machine Vision on Mobile Phone Adaptors

Project Background

Electronics, industrial products will generally be based on the company’s design needs in the product surface printing on the corresponding logo or characters, 

LOGO and other features, but often in the printing process will be missing characters phenomenon, such as incomplete character printing, character printing errors, 

font skew and other defects, will seriously affect the quality of consumer concerns and trust in this product, the electronics industry manufacturers and its strict requirements, 

so manufacturers need a set of machine vision inspection system, through industrial cameras, lenses to detect errors or defects in the character. 

Testing requirements

Detecting the defects of character printing on the charger surface, reading the 2D code on the product and comparing with the product SN code for consistency, 

detecting the accuracy of the length and spacing of the pin feet and the verticality of the pin feet. 

The software to achieve the principle of detection

1. Read and compare

1. image acquisition: the first step is to acquire the image to be recognized.
2. image pre-processing: the pixel point gray value on the image will be set to 0 to 255, that is, the whole image will present the obvious visual effect of only black and white.
3. font analysis: the character to be recognized will be segmented, branch processing.
4. character segmentation: at this time, character positioning, character segmentation, the boundary of the string, and then the overall segmentation of the string respectively (can also be a single segmentation), the segmented characters and then do recognition.
5. character recognition: the current character extracted feature vector with the feature template library for template coarse classification and template fine matching to identify the character. After comparing the recognized characters with the read out QR code to see if they are consistent.

2. Character detection and pin detection

1. image acquisition: the first step will be to collect the image to be identified.
2. parameter setting: the detection parameters of the software are set through the detection requirements of the product.
3. software analysis: the software will analyze the captured image and set parameters and template library to do comparison, to determine the good or bad products. the results are given: the software analyzes the good and bad of the product.