Pattern identification method, parameter learning method and apparatus

A pattern identification method for classifying input image data into a first class in which image data does not include a specific object or a second class in which image data does include the specific object by sequentially executing a combination of a plurality of classification processes,

• wherein at least one of the plurality of classification processes comprises:
• a mapping step of mapping the input image data in an n-dimensional feature space as corresponding points representing respective feature amounts of n partial image data obtained from the input image data, where n is an integer equal to or greater than 2;
• a determination step of determining whether the input image data belongs to the first class or whether the next classification process should be executed for the input image data based on whether or not a first value related to a location of the corresponding points mapped in the n-dimensional feature space in the mapping step is larger than a censoring threshold value;
• a selecting step of selecting a classification process that should be executed next from a plurality of selectable classification processes so as to classify the input image data into a plurality of classifications, based on whether or not a second value related to the location of the corresponding points mapped in the n-dimensional feature space in the mapping step is larger than a branching threshold value in a case where it is determined that the next classification process should be executed for the input image data in the determination step, wherein the classification processes which are not selected are not executed;
• and
• a terminating step of terminating a processing for the input image data in a case where it is determined that the input image data belongs to the first class in the determination step.