PROBLEM TO BE SOLVED: To perform the classification of image data to be adaptable to each user's environment and to provide its classification method with a learning function.;SOLUTION: In this image classification learning device 10, an image acquisition part 11 acquires an image data group for classification from an image DB 17, and a representative image corresponding to a classified class is determined by a representative image determination part 12. An image retrieval part 13a of a case collection part 13 retrieves a similar image using the representative image as a query, and an image data extraction part 13b extracts case data from the acquired image data to be stored in a case data storage part 16. An identification function generation part 14 performs learning based on the extracted case data to calculate an identification function to determine the classification. An identification processing part 15 classifies the image data group into classes using the calculated identification function.;COPYRIGHT: (C)2010,JPO&INPIT
展开▼