基于ARM的类人足球机器人视觉识别系统的设计  

Design of vision recognition system of humanoid soccer robot based on ARM

作　　者：席文平 张健[1] 

机构地区：[1]辽宁工业大学电气工程学院,辽宁锦州121000

出　　处：《微型机与应用》2015年第9期51-53,共3页Microcomputer & Its Applications

摘　　要：针对类人足球机器人在复杂赛场环境中对目标色块识别能力的不足,设计一种基于ARM的类人足球机器人视觉识别系统。软硬件设计上,以搭载Linux系统的三星S3C6410作为主控图像处理芯片,用飞思卡尔MC9S12XS128单片机作为舵机控制器,以罗技C270摄像头作为视觉传感器。分割识别算法上,采用YUV颜色空间下的阈值法与区域生长的混合分割算法对色块信息进行快速分割和识别。实验表明,该设计可准确识别出较复杂环境中的目标色块并定位出中心坐标。Aiming at the shortcomings of the target color recognition ability of the humanoid soccer robot in complex field environment, this paper designed a kind of humanoid soccer robot vision recognition system based on ARM. In terms of hardware and software design, Samsung S3C6410 with Linux is chosen as the main image processing chip, Freescale Mcgs12xs128 microcontroller is applied as the servo controller, and Logitech C270 camera is used as the visual sensor. As for the image recognition algorithm, a mixed segmentation algorithm of the threshold combined with region growing algorithm in the YUV color space is used to segment and recognize the color information rapidly. After verification, this design can accurately identify the color swatch and locate the center coordinate in a complex environment.

关 键 词：类人足球机器人 ARM 单片机 混合分割 图像识别 

分 类 号：TP242.62[自动化与计算机技术—检测技术与自动化装置]