The Hardware Design Of Defect Detection System Based On Embedded Machine Vision

In modern industrial production, visual detection is often indispensable link. For example, the appearance of auto parts, drugs package, quality of IC characters printed on circuit board and welding, etc., all require many workers testing by the naked eye or with microscopes for observation test. Substantial manual testing not only affect the efficiency of factories, but aslo bring many unreliable factors that directly affect product quality and cost. Moreover, many processes are not testing for the detection appearance, and need to obtain accurate testing data, such as parts of the width of the hole diameter, and coordinate of the benchmark, etc., which are very difficult to quickly accomplished by human eyes. With the development of image processing technology and digital technology, a new dectection methods— visual machines comes into being.Machine vision system means capture images through machine vision equipment (image capture devices, including CMOS and CCD two kinds), and then transmit the images to the processing units after digital to analog conversion, abstract character factors according to the distributing of pixels, brightness and color of the images. At last control the work of the equipment by the judement through these character factors.Currently embedded system is a hot technology which is developing rapidly, it is the specific application solutions. Because of the advantage of small size, low price, simple development environment, flexible using, reliable operation, it is now widely used in many fields.There are mang large flat float glass production lines in our country now, and in the production process inevitably produce some defect, such as bubble, impurities, the speed of glass pulling is rapid and real-time response to demand relatively high, manual testing can not be competent, mature solutions are all based on machine vision now. Currently domestic glass defect detection systems are brought in from outside and the prize is high and maintenance is difficulty. Currently only someinternal PC-based of defect detection system is researched in our country and the technology is not very mature, no reports of large-scale applications until now. In this background we combine the advantage of embedded system and machine vision to research and develop an embedded glass defect dectection hardware system. Entire hardware system includes CCD imaging system, image capture, image processing and computer interface, and we devote lots of energy to the reliability of the system to ensure it to stable working in industrial production environment.