Method and apparatus for road lane recognition by surface region and geometry information

PURPOSE: A road recognition method and an apparatus thereof using road region and geometric information are provided to effectively recognize a road or a lane by detecting a road region from an input image using information selected from energy of the input image, contrast, uniformity, and correlation. CONSTITUTION: A road recognition method using road region and geometric information comprises the following steps: obtaining a surface feature standard vector for a region of interest from plural road images in a DB, and comparing the surface feature standard vector with a surface feature vector of an input image to determine a road region; detecting a crossing point with a lane inside the road region using geometric information of the lane, and obtaining a burning-off point for recognizing the boundary line and the lane; and determining a lane change direction, a lane change initiating point, and a lane change finishing point by calculating the location of the lane boundary line, angle, and moving speed, and calculating the estimated time. A road recognition apparatus using road region and geometric information includes a road region determining unit, a lane boundary line and lane detection unit, and a calculation unit of the lane change estimated time. [Reference numerals] (AA) Start; (BB) End; (S011) Setting a region of interest from a specific image in DB including multiple road images according to the same length of distance; (S012) Dividing the region of interest by multiple segmented regions and calculating each surface feature vector about each segmented region; (S013) Calculating the average of each surface feature vector to produce the average surface feature vector about the region of interest; (S014) Calculating a surface feature vector of the segmented regions in the input image and comparing the segmented regions with the surface feature vector included in the region of interest according to the distance matching for the segmented regions to determine a road area