NOVELTY – The media to be analyzed is acquired by resolving the formation of media sampling space. The decomposition of media in the multi-scale is performed. The disturbance of the least significant property of original media is randomized. The regions of interest in the media analyzed are selected. Certain set of statistical descriptors on each generated region is calculated. The descriptors are normalized in relation with the measurement present in original media and collected according to specific criterion. USE – Progressive randomization method for analysis and reasoning of multimedia information such as image, sound and video. ADVANTAGE – The problems caused during multimedia content categorization, hidden message detection, content tampering detection, robust watermarking, recognition of originating portion of media can be solved reliably. DETAILED DESCRIPTION – An INDEPENDENT CLAIM is included for progressive randomization equipment. DESCRIPTION OF DRAWING(S) – The drawing shows an explanatory view of the progressive randomization method by overlapping four regions.