Vision Sensor Simulation And Realization

Visual sensor is the window through which the intelligent robot observing the outer world. Visual sensor is the important part of the intelligent robot. This paper has researched and realized the simulation of lidar and colorful CCD camera. The literature about sensor simulation is deficient, furthermore the literature of lidar simulation is much less. Sensor simulation is the important part of the simulation of intelligent robot. It supplies the data for testing the obstacle avoidance algorithm, navigation algorithm, the communication algorithm and the cooperation algorithm between robots. So the research work of this paper is significative. Lidar simulation has adopted ray tracing algorithm which is used in graphics. The innovation of the algorithm is the use of 2-D grid which records the height of the scene. With the depth image obtained from the simulation of lidar we have do some research on obstacles detection. The result of the test indicates the performance and feasibility of the algorithm. The simulation of CCD camera is implemented through setting the parameters of the view volume in OpenGL and intercepting, saving the pixels of the screen to files. This paper also does some research on virtual camera calibration. Comparing with general camera model, this method acquires the efficiency with losing some precision.