Indoor 3D mapping technique based on 3D image transformation using Kinect sensor

One of the key tasks in robotics is to build a 3D map from unknown indoor environments, which includes rich information of real world environments. This paper presents a new method of a 3D mapping in an indoor environment using Microsoft Kinect which provides a RGB image and Depth information. In this method, FAST features of RGB image are extracted and matched with the current features with the previous one. Estimate geometric transformation algorithm is used to find the best 2D projective and affine transform from the matched pair points of color images. 2D projective and affine transform matrices are reshaped into a 3D transformed one using Kinect’s FOV of RGB camera and depth information of paired points. The new method builds a 3D environment map using the transformed point cloud data from each frame’s RGB image and depth information. For the performance verification, experiments are carried out with a mobile robot equipped a Kinect sensor.