Navigation of Mobile Devices based on Embedded Image Processing with Support of GPS Systems and Sensors
김수용, 카밀 지덱, 블라디슬라브 막심, 이승수 · Journal of the Korean Society of Mechanical Technology · 2013
A recent development in system-on-chip processors provides possibilities of integration performancedemanding tasks, which can be performed only with the help of superior range of computers before. The article introduces advanced control system with embedded image processing to improving GPS localization of autonomous mobile devices in real environment. The main control system is based on ARM (SoC.) architecture integrated in Raspberry PI development board. Presented control systems can handle real-time video capturing, power-efficient image processing operation for example (thresholding, binarization, motion detection, ...etc.) and streaming results images as video. GPS precision can provide only 3 meters precision together with WAAS (Egnos) satellite. Our solution uses GPS solution and embedded image processing to detect border of roads or sidewalks. Our algorithm detects road or sidewalk universally without detection of white line in roadside, because pathways and some roads cannot provide this feature. Control system can be used for small autonomous mobile devices for example parts transport in industrial park of heavy industries, where are long distances between production halls.