五自由度立体视觉平台——目标物体的识别与定位

Five Freedom Stereo Vision——Object Recogniton and Position

从立体视觉与机器人控制集成的角度出发,建立了一个主动立体视觉跟踪和定位系统,用于柔性装配线中装配零件的运动跟踪和装配工位的精确定位。讨论了使用立体视觉进行快速目标识别和定位的方法,该方法将颜色分割与基于形状特征相结合,提出采用基于局部图像的HSV阈值分割和形状识别相结合的图像处理方法,该方法经过HSV阈值和形状判据,准确地识别出物体,得到物体的边界和质心。能快速准确地识别出物体并进行定位,满足柔性生产线小规模定制产品装配的要求。最后给出的实验结果验证了该方法的有效性。

From the view of combination of stereo vision and robot control,an active stereo vision for tracking and location system is built to work on flex line.A method based on local HSV image and shape of object to recognize object is proposed.After the color segmentation,use the knowledge of object＇s shape to recognize balls.It combines the color segmentation and feature recognition based on shape and can recognize object correctly and quickly to satisfy the requirement of assemblage of small scale.The system relies on the two fundamental cubes of color and shape to achieve high rates of detection,tracking and recognition.The program uses HSV（Hue,Saturation and Value） color space to classify object or background.The result at the end of paper testifies the availability of the method