METHOD FOR TRANSFORMING IMAGE AND APPARATUS THEREOF
摘要
A method for converting an image and an apparatus thereof are provided to minimize an increase of overhead and a bit rate generated in a converting process and obtain an encoded image data stream compressed with high efficiency by applying KLT conversion with an optimal compression rate to an object with motion and divided background parts in an inputted image frame. A method for converting an image comprises the following steps of: detecting whether there is motion for an inputted image frame(310); dividing the inputted image frame into a background and a foreground according to the detected motion(320); and converting a block within the inputted image frame into a frequency area by applying different converting algorithms to the divided background and foreground(330). A KLT(Karhunen-Loeve Transform) converting algorithm is applied to the divided background.