嵌入式机器视觉系统的图像采集及显示技术

Image Acquisition and Display Technologies for Embedded Machine Vision System

针对机器视觉应用中图像采集与显示技术不规范、可靠性差的缺点,提出一种基于嵌入式技术的采集与显示方法。引入Video4Linux视频采集应用程序接口编程技术,通过直接读取和内存映射两种方法采集视频图像;同时结合JPEG图像压缩/解压缩库Libjpeg,对JPEG格式编码的图像文件进行解压缩处理;Framebuffer操作设备使用标准化的流程控制LCD的显示输出。该方法实现了采集与显示图像的功能,流程规范、工作可靠,为图像处理算法的开发与验证构建了实验平台。

To overcome the disadvantages of the image acquisition and display technologies in machine vision application, e. g. , lack of stand- ardization and poor reliability, an acquisition and display method based on embedded technology is proposed. In the method, Video4Linux, the video capture application programming interface for Linux is introduced; and the video images are collected by measures of direct readout and memory mapping. Combining with Libjpeg, the JPEG image compression/decompression library, the image files encoded with JPEG format are decompressed. Then by operating Framebuffer equipment, the images are displayed with standardized processes LCD. The method implements the functions of image acquisition and display, the processes are standardized and reliable; and it builds the experimental platform for developing and verifying the image processing algorithm.