Instrument Function Detecting System Based On Image Processing

This paper completely describes the car dashboard instrumentation with depththeoretical analysis; it is a subject close to the actual application. This article covers theinstrumentation system, image processing algorithms, detecting configuration and thewhole implementation processes. Every automobile instrument detection system part iscomprehensively studied.The image processing algorithm is the focus of this article. Image preprocessingalgorithm improves the image processing overall stability; the image calibrationalgorithm can accurately fix the instrument chucking position errors. Then, analysis therelated image processing algorithms partly, including lights, eight digital tubes, pointerangle detect and image matching. The LED color matching, eight digital tubesidentification process, pointer identification methods and targeted image matching is theemphasis. Each test item is for the instrument manufacturer’s actual testing needs. Allthe image processing methods are brought after a lot of practical applications, whichthere are many innovations that improve the effect of instrumentation, and havepractical significance.As a complete visual inspection system, this paper also contains the other parts ofautomobile instrument detection system, including instrument control, camera selection,inspection station-related design and related software. The easy use of R&D projectsoftware is a bright spot, which is reflected in the configuration procedures andconfiguration information, the configuration process has self-learning function on thelight color, the pointer corner detection parameters determinate. The instrumentparameters and detection process is described using configuration files, in a reasonablestructure, which can use a File Editor to modified according to changes in demand. Inorder to improve the efficiency of instrumentation, instrument testing process is highparallelism. Send commands, image acquisition, image processing, result analysis andother processes are independent computing threads. This paper developed a completestrategy for concurrency control. Finally, the overall structures of the software aredescribed, with specific implementation details, including the software, somearchitecture, multi-threaded design and database design concepts.This work is not only a study for those who are working in visual inspection ofinstrumentation and related personnel, but also a typical factory automatic detection applications.