The Research Of Binocular Image Acquisition And Processing System Based On ZYNQ

Image acquisition and processing system is closely related to our life, with the continuous development of science and technology, all kinds of image acquisition and processing system solutions emerge in endlessly, those systems who only have the function of image acquisition and display have already could not satisfy people’s needs, to the present this kind of system function is more and more perfect, for example the application of machine vision system,from the security monitoring to the automatic robot, we can see it, they can replace people to accomplish some impossible tasks, so the study of this kind of system design has the practical significance.We can make use of MCU, DSP and FPGA to implement Image acquisition and processing solutions, but for these solutions, functions they can achieve are relatively single, poor scalability, the shortcomings of applications has great limitations, this project adopts the Xilinx’s "all programmable" So C products ZYNQ platform launched in 2012, this is a single chip implementation in FPGA + ARM architecture of embedded development platform, support collaborative design of hardware and software, ZYNQ development will bring benefit to both of us ARM rich ecological resources, and can benefit from a flexibility and scalability of the FPGA.For this paper, in the base of deeply research of the ZYNQ platform including development mode, I give full play to its advantage of FPGA + ARM architecture and propose binocular image acquisition and processing system overall design scheme. Most tasks of binocular image sensor signal processing are completed in the PL(FPGA), the implementation in the PL project is divided into sub modules, such as image sensor configuration module, signal synchronization module, image format conversion module and image storage module, etc., using the FPGA parallel processing features achieves hardware acceleration. At the same time, in the ARM of the ZYNQ platform I transplant the Linux operating system and Open CV image library files, image display also made improvements on the basis of the traditional, using the transmission speed faster network interfaces, picture to the upper machine(PC) for display, and at the edge of the upper machine can realize the real-time detection and face recognition algorithm.At the end of the paper I give the test results of binocular image acquisition and processing system, the system can collect the two image sensor signals and delieve the real-time image transmission to the PC and it has the ability of real-time processing. Further improve the development of this system it can be applied to the binocular stereo vision system or implementing high-speed image acquisition. This system has the advantages of quick response, high real-time performance and strong scalability, all these advantages prove that this research has certain practical significance.