摘要
首先详细介绍了深度传感器获取深度信息的工作原理,并利用深度传感器获取到用户的深度图像信息。然后,对深度图像进行校正,具体包括深度距离与实际距离的转换、深度图像到空间三维坐标的转换、深度图像到RGB图像的配准。最后,利用最优阈值法实现了将预处理后的深度信息中人体与背景的分割。实验结果表明,此方案可以比较完整的提取出人体区域。
In this paper,the operating principle of the depth information obtained from the Kinect was introduced in detail,and the depth image information of the user was achieved by Kinect.After that,the depth image was adjusted,including the transformation between depth distance and actual distance,the transformation of depth image to 3 D space coordinate,the registration of depth image to RGB image.At last,the people and the background were segmented in the preprocessed depth information by using optimal threshold method.The results of the experiments validated that this scheme could extract the human body completely.
出处
《信息技术》
2014年第1期109-112,共4页
Information Technology
