POSITION RECOGNITION METHOD FOR INTELLIGENT MOBILE ROBOT USING NATURAL LANDMARK, ARTIFICIAL LANDMARK, AND ENCODER

摘要

Provided in the present invention is a position recognition method for an intelligent mobile robot using a natural landmark, an artificial landmark, and an encoder. The intelligent mobile robot comprises: an image obtaining unit which obtains an image; an image processing unit which processes the obtained image; a map generating unit which generates a map using the image processing results; a position recognizing unit which determines the position of a robot using the image processing results and the generated map; and a position combining unit which combines the determined position and the information on an encoder. The position recognition method for an intelligent mobile robot comprises as follows: a step of checking for a natural landmark or an artificial landmark in an inputted image; a step of removing noise on the above image where the natural landmark or the artificial landmark is checked; a step of extracting a feature point of the natural landmark from the image; a step of extracting a feature point of the artificial landmark from the inputted image; a step of performing a binarization algorithm using the extracted feature points; a step of matching by comparing the information on the performed binarization algorithm with the registered natural and artificial landmarks; a step of performing a position correcting motion using a Kalman filter; and a step of combining the corrected position and the encoder information.;COPYRIGHT KIPO 2014;[Reference numerals] (AA) Input image; (BB) Check for a natural landmark inside the image; (CC) Check for an artificial landmark inside the image; (DD) Image preprocessing for removing noise; (EE) Extract a feature point of the natural landmark; (FF) Extract a feature point of the artificial landmark; (GG) Is there an external light source (sunlight, lamp)?; (HH) Apply a general binarization algorithm; (II) Apply a strong binarization algorithm against the external light source; (JJ) Compare and match an object on the current screen with the registered natural landmark and artificial landmark; (KK) Remove data noise using a Kalman filter and correct the position; (LL) Detect a heading angle along a straight line fixated to the object; (MM) Is there an artificial landmark?; (NN) Input encoder information; (OO) Calculate the movement/direction; (PP) Is there a natural landmark?; (QQ) Calculate the position using an encoder moving distance; (RR) Calculate the position by converging the natural landmark and the encoder data; (SS) Output the coordinate; (TT) End