Study On Sound Source Localization For Service Robot Based On Auditory&visual

With the development of intelligent robot technology, service robots engage go in for lots of work instead of human in recent years. In order to get the real-time status of the actual service environments, robots need to gain the real-time processing information of the environment. Target positioning become a key technology for service robots performing tasks. Auditory and visual systems are very important organs of the human perception of the environments, and they are the main research directions for the robot’s information. Visual positioning can visually track the target, but it requires the brightness of light and high openness. Auditory localization can collect omnidirectional sound source, under illumination, visibility, obstacle limitation, but it cannot get the remote acquisition target visual image. This paper focused on the urgent problem that a service robot object orientation, comprehensive auditory and visual target location advantage. For this, the paper mainly includes two aspects as follows:1. Dimensional space target sound source localization based on microphone array of tetrahedronBased on microphone array research, this paper proposed the use of tetrahedron microphone array to locate the sound source. This paper made the analysis and the simulation analysis theory which provide an important theoretical basis for the later design and improvement of service robot auditory system.2. Experimental platform has been built for sound source localization based on auditory- visual trackingThis paper selected the omnidirectional microphone, and constructed tetrahedral array as a collection of sound information device. I sampled synchronously four voice signals through the NI USB, and sent them to the PC platform which designed by LabVIEW. Through the NI-DAQ module program design, this paper achieve the receiving and processing of sound signals. Generalized cross correlation algorithm determines the time delay of microphone and the sound source azimuth angle through the array geometry models. At last, the sound source location information which in the form of instruction will control visual device tilt the stepping motor to rotate the camera, and this will achieve sound source localization tracking.