Method of processing image formation

A method of processing image information for recognizing the movement of an object on the basis of frame images obtained in a three dimensional space in time series manner comprising the steps of:

• entering a stereoscopic frame image of the object;
• extracting edges of the object from the stereoscopic image;
• locating characteristic points on the edges from the stereoscopic image;
• determining local features of the edges from the stereoscopic image;
• on the basis of the local features from the stereoscopic image dividing the edges into a plurality of segments, storing the segment information into a boundary representation model data structure;
• with reference to characteristic points, selecting tracing points on the segments;
• determining a three-dimensional correlation of the tracing points between frames of the tracing points;
• and
• detecting position, attitude and a kinetic parameter, the kinetic parameter consisting of a rotational component and a translational component, of said object on the basis of the three dimensional coordinates of said three-dimensional correlated point.