Automatic Checkout System Design Of LCD Projector

At present, the projector's quality is detected by manual basically in the domestic market. However, due to low efficiency and visual fatigue caused by long time working, it often leads to false detection. Meanwhile, the intense light may do great harm to workers' eyes. Therefore, we need a more effective detecting method to replace manual one. In this dissertation, an automatic inspection system of LCD projector based on machine vision technology was designed and implemented. Through simulation experiments in the laboratory, the feasibility of the system was validated.In order to achieve rapid and reliable detection, the high-speed image acquisition equipment and computer workstations were chosen as the working platform to conduct the quality detection of LCD projector in terms of system hardware. The image acquisition equipment included high-speed digital CCD camera, lens and 1394b card. FL2-14S3M/C-C digital camera, Flea2 series from Point Grey Company, whose maximum resolution is 1392x1032 and acquisition rate could up to 30 pictures per second, was selected. LM12JCM fixed-focus lens from Kowa Company was adopted.Aiming at the requirements of projector quality detection, several Detection Tool modules in terms of detecting system software were designed such as dust examination, Liquid crystal board flaw examination, focusing examination, RGB movement confirmation and synthesis examination through digital image processing algorithms based on Microsoft's Visual C++6.0 development platform. In the practical application, LCD projector was detected to decide the projector's quality by combining these modules.The main functions of system software include detecting system settings, image processing and operation detection. System settings section includes models setting, image detection and camera settings. Image processing section debugs the system when models increase. Operation detection section detects images of the projector, saves image information about non-conformance products automatically and marks out the unqualified area.The results of simulation experiments confirm that the detection system developed in this dissertation possesses practicable hardware and reliable software, which provides a feasible solution to projectors' quality detection.