Abstract: Data compression and decompression methods for compressing and decompressing data based on an actual or expected throughput (bandwidth) of a system. In one embodiment, a controller tracks and monitors the throughput (data storage and retrieval) of a data compression system and generates control signals to enable/disable different compression algorithms when, e.g., a bottleneck occurs so as to increase the throughput and eliminate the bottleneck.

Abstract: Embodiments of a three-dimensional wavelet transform are described.

Abstract: Apparatus and method for encoding zerotrees in a wavelet-based coding technique. The method uses a depth-first pattern for traversing the zerotree, i.e., each branch of the tree, from parent to child to grandchild and so on, is fully traversed before a next branch is traversed. The depth-first tree traversal pattern is used to quantize the coefficients of the tree as well as to assign symbols to the quantized coefficients. The method assigns one of three symbols to each node: ZEROTREE ROOT, VALUED ZEROTREE ROOT, and VALUE. By using three symbols and the efficient tree traversal pattern, the method is substantially more efficient at encoding a zerotree than the prior art. Additionally, this concept is applied to the encoding of “vector” zerotrees.

Abstract: A digitized image is encoded by detecting edges in the image, encoding the position and sharpness of the detected edges, filtering the image by a low-pass filter to generate a low-frequency image, and encoding the low-frequency image. A digitized image encoded in this way is reconstructed by generating a horizontal edge image and a vertical edge image from the encoded edge position and sharpness information, synthesizing a pair of high-frequency images by filtering the horizontal and vertical edge images with an edge synthesis filter, decoding the low-frequency image, and performing an inverse wavelet transform on the decoded low-frequency image and the high-frequency images.

Abstract: The invention relates to an encoding method for the compression of a video sequence. Said method, using a three-dimensional wavelet transform, is based on a hierarchical subband coding process in which the subbands to be encoded are scanned in an order that preserves the initial subband structure of the 3D wavelet transform. According to the invention, a temporal (resp. spatial) scalability is obtained by performing a motion estimation at each temporal resolution level (resp. at the highest spatial resolution level), and only the part of the estimated motion vectors necessary to reconstruct any given temporal (resp. spatial) resolution level is then encoded and put in the bitstream together with the bits encoding the wavelet coefficients at this given temporal (resp. spatial) level, said insertion in the bitstream being done before encoding texture coefficients at the same temporal (resp. spatial) level.

Abstract: An image compression system for implementing a zerotree wavelet compression algorithm. The compression system uses a wavelet based coding system which takes advantage of the correlation between insignificant coefficients at different scales.

Abstract: Data elements, preferably representative of video data, are logically divided into blocks. In a bit-wise fashion, each block is inspected to determine whether the data elements for that block may be represented in a highly compact format. If a given block may not be represented in this manner, it is sub-divided into blocks having smaller dimensions. This process of identifying suitable blocks and sub-dividing is recursively repeated until minimum block dimensions are reached. The same result may be achieved through the use of a plurality of ascending tables that are constructed by repetitively forming tables of reduced data elements. The plurality of ascending tables is traversed and, based on the reduced data elements, blocks of data are identified that are susceptible to the highly compact format. Wavelet transforms are preferably used to provide video data to be compressed.

Abstract: A scalable motion image compression system for a digital motion image signal having an associated transmission rate. The scalable motion image compression system includes a decomposition module for receiving the digital motion image signal, decomposing the digital motion image signal into component parts and sending the components. The decomposition module may further perform color rotation, spatial decomposition and temporal decomposition. The system further includes a compression module for receiving each of the component parts from the decomposition module, compressing the component part, and sending the compressed component part to a memory location. The compression module may perform sub-band wavelet compression and may further include functionality for quantization and entropy encoding.

Abstract: A system and a method for wavelet-based low bitrate video coding is provided. The system and method of this invention processes each frame of a video sequence based on its content. The discrete wavelet transform of each frame is taken separately. If the difference between the contents of a first frame and a second frame is above a threshold, then the wavelet transformed second frame is quantized and lossless encoded. If the difference is less than a threshold, then motion estimation is employed in wavelet domain. Then the difference between the wavelet transformed second frame and the motion compensated wavelet transformed second frame is quantized and lossless encoded. The quantization is applied by using dynamic adjustment. The quantization tables are optimized based on the coded frame characteristics. In addition to the coded frames, the lossless coded motion vectors are also transmitted to the decoder unit.

Abstract: A method for eliminating the flicker of a light source such as a florescent light having the first step of setting a first frame rate for the capture of a series of frames. Then, setting a capture integration period to a first integration period, and capturing a set of frames under an illumination source with a first frequency. Thereafter, determining the first frequency of the illumination by using a Fourier transform (FFT) of the set of captured frames. What is also disclosed is an apparatus for performing the above method having an image sensor and a capture control unit coupled to the image sensor.

Abstract: A system is described that comprises a method and apparatus for further compressing an already compressed data stream, using a mathematical exclusive-OR computation. The method comprises: receiving a data stream that has been compressed using a first compression protocol; decompressing the data stream using the first compression protocol; reducing temporal redundancy in the data stream using an exclusive-OR comparison protocol; and recompressing the data stream using the first compression protocol. The apparatus comprises: a decompression unit to receive and decompress a compressed data stream to produce decompressed data; an exclusive-OR comparison unit, coupled to the decompression unit, to perform a mathematical exclusive-OR operation on the decompressed data to produce resultant data; and a recompression unit to compress the resultant data to produce a further compressed data stream.

Abstract: An apparatus for image coding using tree-structured vector quantization based on a wavelet transform and a method therefor are provided. The apparatus for image coding using a tree-structured vector quantization based on wavelet transform has a wavelet transform unit, a vector construct unit, an error vector unit, a scan unit, a first quantization unit and a second quantization unit. The wavelet transform unit wavelet transforms an input image signal. The vector construct unit constructs vectors, each having a tree structure in a different direction, using the wavelet transformed result. The error vector generation unit generates a plurality of error vectors by setting one of the vectors as a basic vector and performing a calculation on each of the vectors remaining with respect to the basic vector. The scan unit scans the coefficients of each of the basic vector and error vectors in a different direction.

Abstract: A group of video plane (GOV) layers in which the encoding start time is absolute time with an accuracy of one second is provided as a coded bit stream. A GOV layer can be inserted not only at the head of the coded bit stream but at an arbitrary position in the coded bit stream. The display time of each video object plane (VOP) included in the GOV layer is represented by modulo_time_base which represents absolute time in one second units with the encoding start time set as the standard, and VOP_time_increment, which represents in millisecond units, the time that has elapsed since the time point represented by the modulo_time_base.

Abstract: The invention relates to a video coding method applied to a sequence of frames and based on a tridimensional (3D) wavelet decomposition with motion estimation and compensation on couples of frames. The method comprises the steps of organizing the wavelet coefficients resulting from said decomposition in a structure of macroblocks, separated by resolution flags, and blocks with a size fitting the lowest approximation sub-band of the decomposition, scanning the coefficients of each 3D macroblock in a predefined order connected to the spatial orientation and the location of each of its blocks, and encoding said scanned coefficients bitplane by bitplane. Moreover, the encoding step includes a global bitrate allocation sub-step between the macroblocks, by means of a global distortion minimization.

Abstract: In one embodiment, an image is broken up into multiple regions or segments, where each segment may be of arbitrary shape, and a transform (multi-scale or otherwise) is applied on the set of segments. In another embodiment, pattern adaptive prediction is used when predicting the next finer level of the transform pyramid. The pattern adaptive prediction uses the parent grid to determine what geometry of a filter is to be used when predicting the child grid. At the boundaries of the domain, the pattern adaptive prediction can be coupled with the domain adaptive prediction technique.

Abstract: A method and system for producing decoding the transmission of high-resolution images transmitted as a low resolution spatially scalable FGS encoded base layer and at least one enhancement layer is presented. The low resolution received base layer is representative of a downscaled image of the original image. In this manner, a minimum resolution base layer is transmitted and higher resolutions may be obtained and utilized depending on the available bandwidth and the receiving system resolution capability. In one aspect of the invention, the base layer is decoded and a quality enhancement is next applied to the base layer. The combined base layer and quality layer video frames are then upscaled and the upscaled image is combined with a decoded spatial enhancement layer information. The spatial enhancement layer information fills in resolution lacking in the upscaled base layer/quality layer image. Thus, a high resolution image is formed.

Abstract: The quality of digital images recovered from compressed data in an inter-frame redundancy-removing scheme is enhanced using a self-adaptive feedback scheme in an image compression/decompression system so as to include means for the compensation of the distortion component from prior frame compression in subsequent difference frame compression. Each transmitted frame is stored after a full compress/decompress cycle, and difference data (which includes the inverse of the distortion component from compression of the transmitted frame) representing the difference between the stored frame and the incoming new frame is transmitted. Consequently, the quality of static regions in the recovered images may be improved with each subsequent iteration by taking the distortion component in the prior frame into consideration along with the inter-frame motion information. The feedback loop thus forms a self-adaptive iterative cycle.

Abstract: An apparatus and method for video coding is disclosed wherein one frame is divided into a plurality of blocks. The present invention divides an input frame into a plurality of blocks and transforming the blocks to a frequency domain; rearranges the transformed frequency band-based coefficients in the order of priority based on a degree how much information required for image reproduction is included; and coding the information concerning the position and amplitude of the rearranged coefficients in the order of priority based upon significance and outputting the bit streams arranged according to the significance.

Abstract: A method of compressing digital data is disclosed including the steps of transforming the data utilizing a discrete wavelet transform to produce corresponding transformed data; quantizing the transformed data utilizing a variable quantization determined by a corresponding quadtree structure wherein each of the quadtree leaf nodes has an associated quantization factor utilized in the quantizing of the transformed data. Preferably, the quadtree is determined to be an optimum in a rate distortion sense and encoded utilizing a binary prefix notation followed by a list of quantization factors. The method of Lagrange multipliers can be utilized to determine the optimum to a predetermined number of bits per data item. The present invention has particular application to image data or to video data and in particular frame difference data.

Abstract: A motion wavelet transform zero tree codec achieves high compression and is implemented in hardware of modest size and at very low cost. A wavelet transform is combined with a tree walk technique for encoding the resulting wavelet coefficients. A 2-6 wavelet transform is used. Wavelet coefficients from the transform are represented in a pyramid of wavelet coefficients. An array of zero trees are formed from the pyramid to hold the wavelet coefficients, one coefficient to each node. Significance values for each node are calculated to assist with the tree walk and encoding. Each zero tree is traversed to produce an output of encoded bits. Encoded bits are output directly during the tree walk.

Abstract: A method comprising constructing virtual (Discrete Wavelet Transform) DWT sub-bands from an image by performing the DWT and then applying an inverse DWT upon the virtual sub-bands, the result of the inverse DWT representing an up-sampled version of the image. In an alternate embodiment, an apparatus comprising a computer readable medium having instructions which when executed perform constructing virtual (Discrete Wavelet Transform) DWT sub-bands from an image by performing a DWT upon the image, applying an inverse DWT upon the virtual sub-bands, the result of the inverse DWT representing an up-sampled version of the image.

Abstract: An apparatus and a concomitant method is disclosed for recovering or adjusting quantized coefficients by using a nonlinear method. The method operates by fitting the received signal into one of several predefined classes, and adjusting the signal as appropriate to better fit the best suited class.

Abstract: Video data compression apparatus in which successive groups of two or more images are encoded subject to a predetermined limit on the quantity of output encoded data to be generated for each group of images, comprises a trial encoder for encoding at least a subset of images of a current group of images to generate trial-encoded data for each trial-encoded image; a data quantity allocator, responsive to the quantities of trial-encoded data generated by the trial encoder, for allocating respective quantities of output encoded data for each image in the group of images; and a final encoder for encoding each image of the group of images in accordance with the quantities of output encoded data allocated to that image by the data quantity allocator.

Abstract: Data elements, preferably representative of video data, are logically divided into blocks. In a bit-wise fashion, each block is inspected to determine whether the data elements for that block may be represented in a highly compact format. If a given block may not be represented in this manner, it is sub-divided into blocks having smaller dimensions. This process of identifying suitable blocks and sub-dividing is recursively repeated until minimum block dimensions are reached. The same result may be achieved through the use of a plurality of ascending tables that are constructed by repetitively forming tables of reduced data elements. The plurality of ascending tables is traversed and, based on the reduced data elements, blocks of data are identified that are susceptible to the highly compact format. Wavelet transforms are preferably used to provide video data to be compressed.

Abstract: A system and method are disclosed for providing streaming video over a network by compressing video data in a cascaded fashion. Initially, the system compresses video data to a first intermediate data file using a first transform. Next, the system compresses the first intermediate data file to a second intermediate data file using a second transform. The second intermediate data file is then compressed to a streaming video data file using a third transform. The streaming video data is then transmitted over the network, such as the Internet, as streaming video data. In one embodiment, the first intermediate data file is an MPEG-2 data file and the second intermediate data file is an unconstrained MPEG-1 data file. The first and second transforms may be carried out in transparent mode.

Abstract: A scalable video coding scheme having a single motion compensation loop that generates bi-directional predicted frames (B frames) or predicted frames and bi-directional predicted frames and (P and B frames) coded entirely with a scalable codec.

Abstract: A system and method for performing image compression using multi-threshold wavelet coding (MTWC) which utilizes a separate initial quantization threshold for each subband generated by the wavelet transform, which substantially reduces the number of insignificant bits generated during the initial quantization steps. Further, the MTWC system chooses the order in which the subbands are encoded according to a preferred rate-distortion tradeoff in order to enhance the image fidelity. Moreover, the MTWC system utilizes a novel quantization sequence order in order to optimize the amount of error energy reduction in significant and refinement maps generated during the quantization step. Thus, the MTWC system provides a better bit rate-distortion tradeoff and performs faster than existing state-of-the-art wavelet coders.

Abstract: A code length of image coded data is length-fixed in accordance with a code length representing a required criterion. For this purpose, a discrete wavelet transformation unit 110 performs discrete wavelet transformation of image data. A factor quantization unit 112 performs quantization. An entropy coding unit 113 performs entropy coding of a factor quantized value by arithmetic coding. If the code length of quasi image coded data is longer than the code length designated by a designated image coded data code length inputting unit 102, entropy coded values are deleted in the unit of sub-bands so that the code length of the quasi image code data equals the designated code length. If the code length of quasi image coded data is shorter than the designated code length, adjustment bits are added.

Abstract: A video decoder for decompressing video data to recreate highly accurate and detailed video images. The video decoder includes a programmable gate array which receives a serial data stream of compressed video data containing 32 bit data words. The programmable gate array converts the serial data stream of compressed video data to a parallel format comprising four eight bit bytes which are sequentially written into a multiformat video codec. The multiformat video codec is a video decoder optimized for real-time decompression of interlaced digital data. The decompression algorithm for the multiformat video codec is based on the bi-orthogonal (7,9) wavelet transform and implements field independent sub-band coding. The decompressed video data is supplied to an imaging device which recreates video images.

Abstract: A method and apparatus for displaying images. In one embodiment, the method comprises displaying a first image at a first resolution level, identifying a location in the first image, and generating a second image for display at a second resolution level different than the first resolution level in response to user input via a user input mechanism. The second image represents a portion of the first image at a different resolution level, which is dependent on the number of times the user input mechanism is utilized or activated (e.g., the number of mouse clicks made with a mouse). The data reuse may also be used for panning images in display windows.

Abstract: A system and method for compressing a video data stream receives a sequence of video frames, each video frame containing an array of image data representing an image. A spatial transform module performs a spatial decomposition transform on the video frames to generate spatially transformed video frames. A temporal transform module performs a temporal decomposition transform on blocks of the spatially transformed video frames, each block containing a predefined number of the spatially transformed video frames in a sequence corresponding to the sequence of the corresponding video frames. The temporal transform module applies a temporal decomposition transform to at least one low spatial frequency subband of data in the spatially transformed video frames so as to generate temporally transformed video data. A data encoder encodes, for each block of video frames, the temporally transformed video data and the subbands of data, if any, to which the temporal decomposition transform was not applied.

Abstract: A digital camera includes a data capture device and data processing circuitry for processing image data representing captured images. The data processing circuitry processes tiles of the image data in a predefined order to generated processed image data, which is then stored as a data image file. The tiles are nonoverlapping portions of the image data. Each tile of image data is processing by applying a predefined sequence of transform layers to the tile of image data so as to generate successive layers of transform coefficients. In a preferred embodiment, the transform layers are successive applications of a wavelet-like decomposition transform. While each tile is processed, a predefined set of edge transform coefficients from a plurality of the transform layers are saved in memory for use while processing neighboring tiles.

Abstract: The invention relates to an encoding method for the compression of a video sequence including successive frames organized in groups of frames. Each frame is decomposed by means of a three-dimensional (3D) wavelet transform leading to a given number of successive resolution levels. This method is based on the SPIHT algorithm that transforms the original set of picture elements (pixels) of each group of frames into transform coefficients constituting a hierarchical pyramid in which a spatio-temporal orientation tree—in which the roots are formed with the pixels of the approximation subband resulting from the 3D wavelet transform and the offspring of each of these pixels is formed with the pixels of the higher subbands corresponding to the image volume defined by these root pixels—defines the spatio-temporal relationship.

Abstract: A data encoding method including an encoding step of encoding data using wavelet transform and a multiplexing/transmitting step of appending the information on encoding used for decoding on the decoder side to a bitstream of data encoded using the wavelet transform for transmitting the resulting information. This enables switching of filters depending on the picture or the computational capability that can be furnished by an apparatus and desirable initial setting depending on the input picture or bitrate to realize good picture quality.

Abstract: An in-camera two-stage compression implementation is described that reduces the latency between snapshots to a fraction of that otherwise required by other systems that either process complete compression following each snapshot or that incorporate heavy, bulky, and expensive RAM hardware capable of maintaining several raw luminosity records (unprocessed file containing a digital image). In the 1st stage compression the raw luminosity record is quickly, yet partially, compressed to available RAM buffer space to allow a user to expeditiously capture a succeeding image. When the higher-priority processes, the user shooting pictures, and stage one compression subside, a 2nd stage compression, which is slower but more effective, decompresses the earlier partially-compressed images, and re-compresses them for saving in flash memory until they are distributed to a remote platform to be finally converted to the JPEG2000 format.

Abstract: A frame DWT unit makes subband segmentation of input moving image data in units of frames. Also, a field DWT unit makes subband segmentation of input moving image data in units of fields. First subbands obtained by the frame DWT unit undergo an arithmetic process, and a discrimination unit (104) determines based on a first arithmetic value obtained by the arithmetic process if the frame or field DWT unit is applied to the input moving image data.

Abstract: A scaleable video encoder has one or more encoding modes in which at least some, and possibly all, of the motion information used during motion-based predictive encoding of a video stream is excluded from the resulting encoded video bitstream, where a corresponding video decoder is capable of performing its own motion computation to generate its own version of the motion information used to perform motion-based predictive decoding in order to decode the bitstream to generate a decoded video stream. All motion computation, whether at the encoder or the decoder, is preferably performed on decoded data. For example, frames may be encoded as either H, L, or B frames, where H frames are intra-coded at full resolution and L frames are intra-coded at low resolution. The motion information is generated by applying motion computation to decoded L and H frames and used to generate synthesized L frames.

Abstract: Apparatus and method for encoding zerotrees in a wavelet-based coding technique. The method uses a depth-first pattern for traversing the zerotree, i.e., each branch of the tree, from parent to child to grandchild and so on, is fully traversed before a next branch is traversed. The depth-first tree traversal pattern is used to quantize the coefficients of the tree as well as to assign symbols to the quantized coefficients. The method assigns one of three symbols to each node: ZEROTREE ROOT, VALUED ZEROTREE ROOT, and VALUE. By using three symbols and the efficient tree traversal pattern, the method is substantially more efficient at encoding a zerotree than the prior art. Additionally, this concept is applied to the encoding of “vector” zerotrees.

Abstract: In wavelet transform encoding, high-quality encoding is to be realized by enabling picture quality control from one fractional area to another. An input picture 100 is read out in an amount corresponding to a number of lines required for wavelet transform and buffered in a memory unit 6. The input picture then is wavelet transformed in a wavelet transform unit 2 and quantized in a coefficient quantizing unit 3. In quantizing wavelet transform coefficients, the wavelet transform coefficients are multiplied by weighting coefficients from one sub-band to another. The weighting coefficients are determined using the analysis information of a specified block area in a picture, such as motion information and texture fineness information. This enables fine quantization control in terms of a picture block as a unit.

Abstract: A method is disclosed for recovering image memory capacity, in relation to an image which has been encoded using a linear transform according to a layer progressive mode into L layers, L being an integer value greater than unity, the L layers being stored in an image memory having a limited capacity. The method comprises defining a Quality Reduction Factor (700), being a positive integer value, identifying at least one of the L layers corresponding to the Quality Reduction Factor, and discarding (702) said at least one of the L layers in progressive order in accordance with the Quality Reduction Factor, thereby recovering said memory capacity.

Abstract: Disclosed is an input image data compressor which comprises an image recording controller for generating image information on respective input objects from the input image data, generating image information on a moving object using the image information, comparing the previously input image data with the presently input image data, generating new image information on the moving object and compressing the new image information; a storage unit for storing signals or data generated during compression operation of the image recording controller in a corresponding established address; and a compressed image storage unit for storing the compressed image data compressed by the control operation of the image recording controller.

Abstract: An adaptive region-based, multi-scale, motion compensated video compression algorithm design for transmission over hostile communication channels. The algorithm is embodied in a video encoder that extracts spatial information from video frames to create video regions that are then decomposed into sub-bands of different perceptual importance before being compressed and transmitted independently. The system further uses unequal error protection, prioritized transmission and reconstruction to guarantee a minimum spatial and temporal resolution at the receiver. In particular, the region segmented frames bound both spatial and temporal error propagation within frames. A connection-level inter-region statistical multiplexing scheme is also employed to ensure optimal utilization of reserved transmission bandwidth.

Abstract: According to the present invention, an image transmitter apparatus which has a data structure suited for receiving a request for diagnosis over images and a received image viewing apparatus arranged for displaying desired images a received side are provided.

Abstract: A digitized image is encoded by detecting edges in the image, encoding the position and sharpness of the detected edges, filtering the image by a low-pass filter to generate a low-frequency image, and encoding the low-frequency image. A digitized image encoded in this way is reconstructed by generating a horizontal edge image and a vertical edge image from the encoded edge position and sharpness information, synthesizing a pair of high-frequency images by filtering the horizontal and vertical edge images with an edge synthesis filter, decoding the low-frequency image, and performing an inverse wavelet transform on the decoded low-frequency image and the high-frequency images.

Abstract: A wavelet transform system and an inverse wavelet transform system are disclosed that respectively implement a wavelet transform and an inverse wavelet transform. Semi-orthogonal standard wavelets are used as the basic wavelets in the wavelet transform and the inverse wavelet transform. As a result, two finite sequences of decomposition coefficients are used for decomposition in the wavelet transform. Furthermore, two finite sequences of reconstruction coefficients that are derived from the two finite sequences of decomposition coefficients are used for reconstruction in the inverse wavelet transform. The finite sequences of decomposition and reconstruction coefficients are not infinite sequences of coefficients that have been truncated. Furthermore, in one embodiment, downsampling is not used in the wavelet transform and upsampling is not used in the inverse wavelet transform.

Abstract: A method for motion compensation adaptive image processing, which processes an image data received from an external source, stores the processed image data and restores the stored data to the received image data, the method comprising the steps of analyzing the received image data into image data having relatively high frequency component and relatively low frequency component, compressing/coding the image data having relatively high frequency component and relatively low frequency component by allocating predetermined bits, dividing the compressed/coded image data into a value corresponding to relatively high frequency component and a value corresponding to relatively low frequency component, and decoding the values and restoring the received image data based on the decoded image data

Abstract: An apparatus and a concomitant method is disclosed for encoding wavelet trees to generate bitstreams with flexible degrees of spatial, quality and complexity scalabilities. The zerotree entropy (ZTE) encoding method is extended to achieve a fully scalable coding method by implementing a multiscale zerotree coding scheme.

Abstract: An information insertion/detection system for detecting and inserting DCT coefficients in image signals includes a resolution analyzer for receiving image code data and outputting the image code data and image size information of the image code data, an insertion pattern determining unit for receiving the image code data and the image size information and outputting the image code data and pattern information, an information inserter for receiving the image code data and the pattern information and producing image code data by inserting information of the DCT coefficients into the image code data and a video analyzer for receiving the image code data and producing information of insertion strength of the DCT coefficients. The information is inserted in DCT coefficients so that the information can be detected by referring to an image size of MPEG data inputted to the system when the image data is converted into a reference image size.

Abstract: A method and apparatus for performing video compression and decompression using dynamic quantization and/or encoding. According to one aspect of the compression method, an image is recursively filtered into its constituent components each represented by a matrix of coefficients. Next, the level of correlation of a first of the matrices of coefficients is determined. Based on the level of correlation of this first coefficient matrix, one of a first quantization technique and a second quantization technique is selected. The first coefficient matrix is then quantized using the selected quantization technique. According to another aspect of the compression method, an image is digitally filtered into its constituent components, each represented by a matrix of coefficients. At least certain matrices are quantized to generated a first quantized matrix.

Abstract: A color transform unit transforms RGB components constructing image data into C1 to C3 components. Discrete wavelet transform units perform wavelet transform on the C1 to C3 components and generates coefficients of different number of subbands for at least two components. Coding units stepwisely perform reversible transform on the coefficients in a subband unit, and store the code data into buffers. A fixed-length data reading unit reads the code data stored in the buffers by fixed lengths, and a code output unit outputs code data where the read code data are arrayed.