Design Of Smart Camera Application Software And Implementation Of GigE Vision Protocol

With the rapid development of microelectronics and integrated circuit technology,and the increasing production efficiency and higher precision requirements of industrial manufacturing,embedded machine vision systems are becoming more and more widely used.The Smart Camera is a highly integrated micro-machine vision system that combines image acquisition,image processing,communication and display in a single camera,providing a versatile,modular and highly reliable machine vision solution.Therefore,the development of image processing applications and communication functions for smart cameras is of great significance for smart cameras.The smart camera development platform and overall structure are first introduced,then designs the smart camera software architecture based on the software and hardware synergy idea,and realizes the video frame queue management under the V4L2 framework.Then the image processing module function is analyzed,and the QR code recognition function is realized.The realization process of the QR code recognition function includes: image binarization hardware acceleration based on HLS,Zbar transplantation of third-party image processing library and QR code implementation based on Zbar Code identification.When implementing the smart camera communication module,the GigE Vision 2.0 protocol is analyzed in detail,including device discovery mechanism,GVCP and GVSP frame structure,GVCP and GVSP communication mechanism.Finally,the GigE Vision protocol is implemented on the Zynq platform,including the main functions of the device discovery mechanism,the GVCP control sub-protocol,and the GVSP streaming sub-protocol.The video frame buffer queue management function is tested by using the Linux kernel Debug function.The QR code recognition function of the smart camera is verified by the actual QR code image.The results show that the software architecture runs normally and the QR code can be correctly identified.The Wireshark and task manager are used,and the GigE Vision protocol implementation code is tested in conjunction with the PC-side MFC program to perform device discovery,packet loss retransmission,and image transfer.