Moving Object Detection on the Side of Vehicle using Camera and LRF Sensor Fusion in Urban Environment

In this paper, we present an advanced moving object detection algorithm in urban environment. Our system is designed to reconstruct a three-dimensional world model. In this reconstruction process, we have to separate the ground structures that is static objects and moving objects when the car is running through the town to make the accurate 3-D world model including ground structures only. The system consists of two vision cameras and two laser scanners to capture lateral data and GPS, wheel encoder, and IMU to get the ego-motion of the vehicle. To detect moving objects on the side of vehicle, we use depth information and image data provided by laser scanners and cameras. In this algorithm, it is important to calculate the ego-motion of the vehicle using internal sensors. Finally, we can reconstruct the 3-D world model that is composed of the static objects by using the accurate moving object detection algorithm.