An investigation of the use of Kinect sensor for indoor navigation

Robot navigation in an unknown and dynamic environment has been a challenging research not only among researchers in the area of Robotics, but also in many learning institutions. The prevailing school of thought is to create more complex sensors to feed the robot with a better vision. However, the trade-off for more complex sensors is cost. Unfortunately cost constrain is unforgiving for university researchers and robotics enthusiasts. This paper reports on an undertaking by the authors on the implementation of robotic vision and navigation system using Kinect sensor making it a low cost system for research purposes in the area of robot navigation. For this system, the point cloud to laser scan function was used to convert the point cloud data to laser scan. The odometry and laser data were then passed on to the SLAM Gmapping function to carry out mapping of the surrounding. Once the map was generated, the navigation stack will plan and send command to achieve autonomous navigation.