Research On FPGA-based Machine Vision System For Image Acquisition Hardware

Machine vision image acquisition system, widely used in automated production andproduct quality detection, is an important part of machine vision technology. Not onlycan it release labors from a bald, highly repeated and even eyesight damaging environment,but also can improve productivity, ensure product quality and reduce production costs.Now in China, there are few researches on machine vision image capture, the keytechnologies of which are mainly depended on imports. Therefore, this issue has a practicalresearch value.As eyes for machine vision system, machine vision image acquisition system canprocess images, extract basic features of the measured object and then send the informationto control terminal of the machine vision system or directly to the output display. Basedon FPGA, this paper mainly completed the design and implementation of machine visionimage acquisition hardware system. First of all, a large number of relevant technical data isstudied, and then the basic architecture, with FPGA as main processor and CMOS as imagesensor, is determined. Then, according to the result of an analysis on system resourcerequirements, the selection of the chip is confirmed. Finally, the circuit principle diagramand part of the PCB design are completed in strict accordance with the design standards ofelectronic circuit. In the process of the system design, a combination of signal integrityand anti-electromagnetic interference can make the system work more stably and reliably.Two pieces of high-speed DDR2are used as image data cache, solving the problem of thereal-time transmission and display for high-resolution images effectively. A circuit boardpractical test shows that this system can realize the functions of real-time acquisition, trans mission and display for high-resolution images.