IMAGE PARTITIONING METHOD USING HIGH-ORDER CORRELATION CLUSTERING, SYSTEM PROCESSING THE METHOD AND RECORDING MEDIUM

摘要

PURPOSE: a kind of to utilize high-order relevant cluster dividing method, the identical recording medium of system processing, which provides, realizes that image division quickly move through the broader region of processing. ;CONSTITUTION: input picture is divided into super-pixel (S210) by system. System building hypergraph considers (S220) of specified conditions from two or more adjacent super pixels in the super-pixel obtained by connection. The system extracts each edge of feature vector, and super graph forms connector characteristic pattern, and the super graph of subregion building passes through high-order related clusters (S230). ;The 2013 of copyright KIPO submissions;[Reference numerals] (AA) starts; (BB) terminate; (S210) input picture is divided into super-pixel by system; (S220) system constructs hypergraph and considers specified conditions by connection in the super-pixel obtained from two or more adjacent super pixels; (S230) each edge super graph that system extracts feature vector forms connector characteristic pattern, and the super graph of subregion building passes through high-order relevant cluster