| In recent years, the machine vision-based detection technique has grown into a new science. It was used more and more widely in industry detection. It can improve the speed and reliability of product detection significantly and reduce the production costs. Firstly, the detection system need to get the workpiece image, then extracts the feature after image pre-processing, finally classifies the qualified and unqualified products through comparing to the template image. This paper has made study deeply according to the characteristics of the workpiece online detection system.In this paper, we took screws as an example, discussed the application about machine vision and template matching in workpiece detection. This system MFC, and combined image processing technology to built an experimental system under the VC++6.0. The major tasks of this article are:1. Analysis and study the working principle and the hardware and software composition of the screw-line detection system deeply.2. This paper designed the image processing algorithms which used in detection according to the characteristics of the piece images, including image enhancement, threshold segmentation, edge detection and template matching, etc.3. Extract the orient codes and the edge moment invariants as image feature. Improved algorithm to achieve template matching.The experimental results indicate that the experimental system can select pieces accurately and improve the speed and reliability of detection. This system also can be applied to other areas of image recognition. |
PDF Full Text Request