METHOD AND APPARATUS FOR RECOGNITION OBJECT IMAGE USING GRAVITY

摘要

The present invention relates to a method for image object recognition using gravity. The method for image object recognition using gravity comprises the following steps of: receiving image data to which an object for each frame is inputted; calculating a coordinate space and a vector from the received image data for each frame; extracting image data having a certain rotational component by using a linear transformation based on the calculated coordinate space and vector of the image; extracting a feature vector based on the extracted image data and converting the image data by applying the feature vector to the image data; and recognizing an object from the converted image data. [Reference numerals] (210) Input image data to which an object for each frame is inputted; (212) Calculation a coordinate space and a vector of an image; (214) Extract image data having a rotational component by using a linear transformation; (216) Extract a feature vector; (218) Convert the image data; (220) Object recognition; (222) Object recognition calculation; (AA) Start; (BB) End