Kinect ros 避障

安装kinect v2 for ros

• 安装 iai_kinect2:
  http://blog.csdn.net/sunbibei/article/details/51594824

• 安装时，catkin_make 出错解决方案:

  1. http://blog.csdn.net/Lottie_yx/article/details/78360852
  2. kinect2_bridge hangs and prints “waiting for clients to connect”?
     This is the normal behavior. ‘kinect2_bridge’ will only process data when clients are connected (ROS nodes listening to at least one of the topics). This saves CPU and GPU resources. As soon as you start the kinect_viewer or rostopic hz on one of the topics, processing should start.

• catkin_make 编译完成后，下面第一条指令是把kinect打开，至于要显示什么数据，你还需要执行后两条指令获得相应的图片

    # start the kinect
    roslaunch kinect2_bridge kinect2_bridge.launch

    # 查看点云
    rosrun kinect2_viewer kinect2_viewer kinect2 sd cloud

    # 查看图像
    rosrun kinect2_viewer kinect2_viewer kinect2 sd image

kincet特性

• RGB 和深度图片的分辨率为640*480, 30帧每秒
• 深度相机的有效精度范围是0.4-3.5米
  
  

避障：

Real-Time Obstacle Detection System in Indoor Environment for the Visually Impaired Using Microsoft Kinect Sensor

2016 sci-Hindawi

解读：

• 用RGB相机和深度相机获取点云数据，矫正参数:N. Burrus, “Calibrating the depth and color camera,” 2014
• 用Voxel Grid Filter 方法下采样 the Point Cloud 点云数据，去除人两侧75厘米外的点
• 采用Random Sample Consensus (RANSAC) algorithm用于点云数据分割
• 分割出地面后，每个障碍物点云簇中点与点之间的距离小于一个门限，Euclidean Cluster Extraction，PCL库中的类会提供一些障碍物的大小
• 门检测流程，实验表明，在摄像机离门趋近45度角时效果更好
  
  实现： c++, PCL, PCL 支持OpenNI 3D 接口，支持kinect数据