摘要
为了满足现实生活中对物体体积实时测量的需求,提出了一套基于Kinect 2.0深度图像处理的快速体积测量方案。首先,使用Kinect 2.0深度传感器获得深度图像及对应的点云数据矩阵,并对深度图像进行增强、二值化、目标提取等预处理,定位出目标物体。然后,通过像素点的统计和点云数据的处理得到目标物体的面积、高度。最后由面积和高度完成目标物体的体积计算。经验证这种方法的体积测量误差控制在3%以内,完全满足实时性的误差要求,又由于深度图像较彩色图像不易受光线、背景的干扰,使得该方法具有很强的鲁棒性。
In order to meet the demand of real-time volume measurement,a fast volumetric measurement scheme based on Kinect2. 0 depth image processing is proposed. Firstly,Kinect 2. 0 depth sensor is used to obtain depth images and the corresponding point cloud data matrix,and the depth image enhancement,binarization,target extraction and other pretreatment methods are used to detect the location of the target object.Then the area and the height of the target object are calculated through doing statistics of pixels and processing the point cloud data. Finally,the volume of the target object is calculated via area and height. It is proved that the volume measurement error of this method is controlled within 3%,which fully meets the error requirement,and because the depth image is less susceptible to light and background than color image,so that this method has a strong robustness.
出处
《微型机与应用》
2017年第7期35-38,42,共5页
Microcomputer & Its Applications