基于Kinect的典型零部件识别与定位

针对自动化拆卸的零部件识别与定位问题,提出了一种结合深度信息的典型零部件图像识别与定位方法.利用Kinect传感器获取彩色图像与深度图像,提取出两者之间的仿射变换矩阵,实现彩色图像的矫正;根据相关系数匹配法实现矫正后的彩色图像零部件识别;利用彩色图像与深度图像的对应关系对零部件进行定位.针对典型零部件,对Kinect传感器获取的图像进行仿真实验与处理,结果表明,该方法能对目标进行识别与定位,验证了方法的有效性.

Identification and location of typical components based on Kinect

Aiming at the problem of component identification and location in automatic disassembly process, a typical component image identification and location method in combination with the depth information was proposed. The color image and depth image were acquired with the Kinect sensor, and the affine transformation matrix was extracted so as to realize the correction of color image. In addition, the component identification with the corrected color image was realized with the correlation coefficient matching method, and the components were located with the corresponding relationship between color and depth images. Aiming at the typical components, the simulation experiment and processing for the image obtained with the Kinect sensor were carried out. The results show that the proposed method can identify and locate the target. And the effectiveness of the method is verified.