An Inspection System For LCD Projectors Based On Machine Vision

In the current socialized production, the artificial detecting method, which is restricted by different conditions, is usually used in the product test of the assembly line. Based on the machinery vision, the development of industrial image inspection system presents feasible schemes to solve the problem. The machinery detection reduces load intensity on workers and ensures high accuracy and reliability of test results.This article designs a set of detection system for LCD projector based on machinery vision technology. Firstly, the signal generator sends a standard image signal to the projector. After projection, some images are captured at an appropriate location by an industrial camera and sent to a PC workstation by image acquisition card. And then the workstation will process the images correspondingly by image processing algorithm to satisfy the operation needs for next step, finally it will compare the processed images with the feature of original standard images and finish the evaluation work of quality.During the design of hardware for the system, the industrial PC and high-speed image acquisition device are employed as the working platform. And the image acquisition device includes a digital camera, len and an 1394b card. The FL2-14S3M/C-C camera made by POINT GREY CO., has a maximum resolution of 1392×1032, and the LM12JCM-type len of Kowa is used.The Visual C++ 6.0 of Microsoft is used as the software platform of testing system, which enables fast and efficient implementation of algorithms, but also increases the portability. The system includes dust detection, RGB color detection, focus detection and image distortion detection modules, and the software functions involve three parts of system setup, image processing and operation testing. For unqualified products, the system will automatically save the information of unqualified products.The paper adopts proper hardware devices and establishes the simulation laboratory. The software keeps stable running during the period of detection. The experimental results show that the detection system developed in this article is capable of replacing labor and achieves the automatic detection of projector.