An Obstacle Detection System Using a Light Stripe Identification Based Method Janne Haverinen, Juha Röning Infotech Oulu and Department of Electrical Engineering University of Oulu FIN-90570 Oulu, Finland email: johannes@ee.oulu.fi, jjr@ee.oulu.fi Abstract A light stripe tracking and identification method is pro- posed for a structured light based obstacle detection sys- tem operating in an outdoor environment. The method makes the structured light based detection system more ro- bust and applicable to use outdoors as aid for navigation. The method differentiates between the structured light pro- duced by a light stripe projector and the light stripe kind patterns caused by ambient illumination. The centre of gravity of the segmented light stripe is tracked by using a Kalman filter. The position information together with the other properties of the stripe segment, including intensity, length and orientation are used to identify the same light stripe segment in adjacent images. By using a pulsed light source it is possible to differentiate between true and false light stripes depending on their time of appearance. Dur- ing the project, a working obstacle detection system for a partly structured outdoor environment was implemented. 1. Introduction There are several approaches to sensing the environ- ment of a robotic vehicle. The typical methods have uti- lized a radar principle or passive computer vision methods, such as stereo vision [1],[6],[8],[4],[3]. Depending on the application, these methods are feasible when the environ- ment is suitable and the speed of the vehicle is low. When the speed of the vehicle increases and the environment is more demanding, more efficiency and reliability are re- quired from the sensory system.With active sensing, the calibration problems and computational costs inherent in passive techniques, such as stereo, can be eliminated. The advantages of using a light stripe based method for environment sensing are that the resulting method can be fast, accurate and simple. The range of multiple points on the target surface can be measured simultaneously and quickly. In good circumstances, structured light can be used to extract 3-D information from the target [2]. Diffi- culties might arise in exceptional environment circum- stances such as snowfall, fog, bright sunshine or uneven ground. A ranging system based on structured light is applica- ble to detecting an obstacle, and can be reliable in favorable circumstances When the circumstances are more difficult, disturbances arise which affect the operation of the system. If these disturbances cannot be eliminated, they may cause false inferences about the environment. The disturbing effects of ambient illumination consti- tute one of the main problems in using a structured lighting based range measurement system in an outdoor environ- ment. This paper addresses the applicability of the light stripe tracking and identification based disturbance filter- ing method for structured lighting based obstacle detection system working in partly structured outdoor environment such as harbor areas. In this method, the light stripe kind patterns are ex- tracted in real time from a sequence of sensor images. The patterns are segmented and traced while they move in sen- sor images during vehicle motion. By pulsing the projected light pattern, it is possible to extract those patterns which appear only when the projector is on and, respectively, those which do not follow the status of the light source. This enables differentiation between the projected struc- tured light and disturbances caused by ambient light sourc- es. 2. Perception system The operating principle of a structured light based ob- stacle avoidance system is presented in Fig. 1 [5]. The main components of the system are the sensor and the light stripe projector. When the working machine moves forward, the sheet of light sweeps the ground ahead of the machine and possible obstacles lying there. The sensor detects the light- profile reflected from the obstacle.