International Business Machines Corporation

Mikihiro Ioka

Junji Maeda

Tomio Echigo

Jung-Kook Hong

Masato Kurokawa

Edge information sensitive to a visual characteristic is efficiently stored, block artifacts are reduced, and highly efficient compression is accomplished by an image compressing method. The system encodes a digital image and includes: an image input for inputting the digital image; a segmenter for segmenting the digital image into a plurality of primitive regions and computing parameters about the luminance and chrominances of the primitive region for each the primitive region; a first merger for merging the plurality of primitive regions to generate first-order block candidates and classifying each of the first-order block candidates into any of a plurality of predetermined patterns; a first clusterer for clustering, among the first-order block candidates belonging to the same classification, the first-order block candidates, where the parameters about the luminance and chrominances of the primitive regions thereof can be approximated with linear transformation, as a first-order block, and representing a transformation coefficient of the linear transformation with a parameter; a second merger for merging a plurality of the first-order blocks to generate second-order block candidates and classifying the second-order block candidate in accordance with the pattern of each the first-order blocks of the second-order block candidate; a second clusterer for clustering, among the second-order block candidates belonging to the same classification, the second-order block candidates, where the transformation coefficients of the first-order blocks thereof can be approximated with linear transformation, as a second-order block, and representing a transformation coefficient of the linear transformation with a parameter; a controller for recursively executing the clustering of the block candidates while raising the order of the block in sequence until the clustering of the blocks becomes impossible; and an encoder for encoding the parameters of the coexisting multi-order blocks.

Method of and system for encoding digital images

An encoding method is provided with which users can select picture quality and a quantity of data in multiple stages and with which an image of higher picture quality can be regenerated by scalable selection, i.e., by further adding data to compressed data that can be decoded. Image data is compressed using: means (62) for segmenting an original image into a plurality of object regions where each region pixels all correlate with one another and for determining a hierarchical structure of the object regions; means (63) for approximating each of the object regions with at least one polygonal surface so that errors of a intensity of luminance and chrominances in each the pixel are less than a predetermined threshold value; means (64) for obtaining residual images by subtracting the approximated image from the original image or by subtracting a decompressed image of a compressed nth-order residual image from the original nth-order residual image; means (65) for compressing the nth-order residual image (n.gtoreq.1) by encoding; means (66) for storing the approximated image and the coded nth-order residual image; and means (67) for decompressing the compressed nth-order residual image from the nth-order residual image.

System for and method of processing digital images

A user may input a motion as a query condition to a motion picture database to retrieve one or more motion picture segments (i.e., motion picture scenes) in which an object moves with a matching motion. A motion information database is built of motion vector sequences extracted from stored motion pictures. A motion is inputted as a query by moving a pointing device, such as a mouse, in a desired motion sequence, rather than as a query expressed in words or as a formula. The inputted motion sequence of the pointing device is converted to a motion vector sequence and compared with stored motion vector sequences to retrieve one or more scenes in which an object moves with a matching motion.

Picture query system using abstract exemplary motions of a pointing

In order to supply data in which scene switching in a motion picture can be easily detected, pixel data 50 of a motion picture is input and stored in the frame buffer 18 as pixel values 52, and the statistical processor 14 operates the mean value .mu. and standard deviation .alpha. of luminous values using all pixel data 50 of one frame. The conversion table processor 16 sets a pixel conversion table expressing the correspondence between pixel values 52 and normalized output pixel values 54 using the mean value .mu. and standard deviation .alpha., reads pixel values 54 sequentially, and outputs output pixel values 54 for one frame to the detector 12 using the pixel conversion table. The detector 12 obtains differential values between previous output pixel values 54 and current output pixel values 54, judges that the number of pixels exceeding the threshold value, Vth, corresponds to a frame where a scene is switched, and outputs the scene change detection signal 56.

Growing community of inventors

Tokyo, Japan

Mikihiro Ioka