AUTONOMOUS TRAVELING VEHICLE, AND OBJECT RECOGNITION METHOD IN AUTONOMOUS TRAVELING VEHICLE

摘要

PROBLEM TO BE SOLVED: To highly maintain recognizing accuracy, and to perform high speed processing in object recognition processing in an autonomous traveling vehicle.SOLUTION: An unmanned forklift includes: a main body part; a model data storage part 46; a camera; a search area determination part 42; an image feature point detection part 43; a featured value calculation part 44; and a matching part 45. The model data storage part 46 stores model data related to the geometric feature of a pallet. The camera photographs the periphery of the main body part at a target position to acquire image data. The search area determination part 42 determines a search area in a predetermined range. The image feature point detection part 43 detects the feature point of image data with respect to the search area. The featured value calculation part 44 calculates the featured value of a matching candidate point extracted from the feature point. The matching part 45 collates the featured value of the matching candidate point with the model data to recognize the position of the pallet on the image data.