Research On The Light Emitting Diode Detection System Of The Positive And Negative Based On Machine Vision

With the development of machine vision, the application of deep into the fields of medical, security, aerospace, military, agriculture, but also is widely used in the detection system of electronic components industry, to ensure product quality, control the production process and the perception of the environment. The process of large-scale industrial production of light-emitting diode, high efficiency, high accuracy and automatic detection system plays an important role in improving product quality, efficiency, cost saving. Because in the process of production, after the encapsulation of diode light often arise about the phenomenon of inversion, because the light emitting diode is the cathode of the pin length is different, and in the process of cutting, the need to judge it is negative. At present most of the factories use artificial detection, the method of artificial cost is high, low detection rate.Therefore, the automatic detection system to develop a fast detection speed, high accuracy and good robustness is very necessary.This paper completed the design of a machine vision based real-time detection system of light emitting diode anode. The first application of CMOS industrial camera work, and then uses the Visual Studio 2010 platform will capture the image preprocessing, detection, judgment with OpenCV software, and finally through the real-time detection system does not conform to the requirements of the work piece.The research in image pre-processing, first gray scale of diode image, then compared the characteristics of three kinds of common filtering methods and the advantages and disadvantages of the Gauss filter algorithm, the collected light emitting diode image filtering options.The method is compared with other methods based on the elimination of noise, can ensure the integrity of information.Then the histogram equalization method for image enhancement, and by comparison with the Sobel, Log, Canny operator, the intersecting cortical model edge extraction algorithm, this method has clear image edge extraction, the loss of information etc. Finally, in the aspect of image segmentation using the Otsu image segmentation algorithm can automatically select the threshold, it can automatically select the threshold value of the two, in the premise of simple calculation, can not be affected by the image brightness and contrast. Image preprocessing method, which laid the foundation for the light emitting diode anode accurate,real-time,efficient detection. In the aspect of detection scheme design, put forward the pixel line processing method, the actual operation of the project mainly.In the aspect of detection scheme design, put forward the pixel line processing method, the actual operation of the project mainly includes the selected line, work, ROI regional setting, setting and programming method. And with the commonly used template matching method is compared, the tests results show that the method has high detection accuracy rate, fast, more suitable for actual production application.Light emitting diode system mainly uses the VC++language automatic detection,this system has the white balance adjustment,the template matching method, the pixel line detection, image capture and stop function, the experimental results show that the system can real-time, fast, accurate detection of light emitting diode.The detection system is committed to achieving full automation production line, not only can improve the production efficiency, alleviate the work burden and workers, to ensure product quality, cost saving, reduce industrial waste.