METHOD OF EXTRACTING CONTOUR IN MOVING IMAGE CODING/DECODING PROCESS

摘要

PURPOSE: A method of extracting a contour in moving image coding/decoding process is proposed to eliminate an area labeling operation requiring a long time, thereby increasing a processing efficiency. CONSTITUTION: The method of extracting a contour in moving image coding/decoding process comprises the steps: performing a rester scanning in a direction from a left upper side of a binary mask to a right lower side thereof until detecting a pixel satisfying a condition 1 or detecting the right lower side of the mask and selecting the pixel as a starting contour pixel when detecting the pixel satisfying the condition 1 and changing an existing object value into an object boundary value and ending the process when detecting the right lower side of the mask; in case the starting contour pixel is selected, rotating in a counter-clockwise until detecting a pixel having the object value and changing a background value of the pixel into a background boundary value and extracting the pixel having the object value as a next contour pixel and also changing the existing object value into an object boundary value. In the condition 1, a pixel value is the object value and also a left pixel value is the background value, the pixel value is the object boundary value and also the left pixel value is the background boundary value, or the pixel value is the object boundary value and also the left pixel value is the background value.