An Embedded System for Vision-based Driving Environment Perception

Active safety guard in vehicles has become a necessity for the ever increasing transportation density. We developed an in-vehicle embedded system for vision-based driving environment perception and hazardous situation warning using the image processing and computer vision technology. Our system runs on the TI TMS320DM642 DSP, which is a high-performance digital signal processor produced by Texas Instruments Corporation. The embedded driving environment perception algorithm consists of two parts: the lane markings detection part and the preceding vehicles recognition part. The system has been evaluated in series of actual driver assistance tests. Herein, the experimental results indicated high vehicle detection rate, low false alarm rate and satisfying robustness