Abstract: A method analyzes a high-level syntax and structure of a continuous compressed video according to a plurality of states. First, a set of hidden Markov models for each of the states is trained with a training video segmented into known states. Then, a set of domain specific features are extracted from a fixed-length sliding window of the continuous compressed video, and a set of maximum likelihoods is determined for each set of domain specific features using the sets of trained hidden Markov models. Finally, dynamic programming is applied to each set of maximum likelihoods to determine a specific state for each fixed-length sliding window of frames of the compressed video.

Abstract: A method and apparatus for preparing an image for downloading over a link. The method includes receiving a user selection for an image to prepare, retrieving current user settings reflective of desired settings for compressing the image and automatically presenting a plurality of variations of the image to the user where each variation is derived using compression settings that are scaled from the current user settings.

Abstract: A hidden watermark and methods of detecting and embedding such watermark is provided. The watermark is resistant to efforts to avoid the message thereof.

Abstract: An image compression apparatus has a connected pixel extraction unit for extracting drawn connected pixels contained in the color document image, a foreground/background image classification unit for classifying connected pixel groups extracted by the compressed connected pixel extraction unit as a foreground image or a background image, a color-by-color-based binary image generation unit for generating binary image data for each color from information on the color of the connected pixel groups classified as a foreground image, a foreground image compression unit for compressing the color binary image data generated by the compressed color-by-color-based binary image generation unit, a background image generation unit for generating a background image, a background image compression unit for compressing the background image generated by the compressed background image generation unit, and a compressed image data generation unit for combining the foreground image data compressed by the foreground image compres

Abstract: In a method and a device for transmission of S+P transform coded digitized images a mask is calculated by means of which a region of interest (ROI) can be transmitted lossless whereby the ROI can be transmitted and received lossless and still maintaining a good compression ratio for the image as a whole. This is possible since no or very few bits can be used for the remaining part of the image. The calculated mask can also be used for transmitting the coefficients needed for a lossless region of interest during any stage of the transmission.

Abstract: A method for automatically determining the region of interest from an image comprises the three steps of (a) analyzing content of the transformed coefficients of an image after a discrete signal transformation and partitioning an image into the interested region and background region that are based on p×p-pixel sub-blocks for performing classification; (b) locating the central point of the interested sub-blocks; and (c) integrating a plurality of the interested sub-blocks from said central point to the boundary of an image to generate a closed and continued region of interest by using a image processing technique with considering the bit-rate requirement.

Abstract: The present invention relates to an article inspection apparatus. The apparatus includes conveying means (6) for conveying an article (5) through an inspection site (I) and rotating the article (5) about an axis (F). The article (5) is viewed by an imaging system including a first and a second image capture means (1A, 1B) and a mirror arrangement (7, 8) that defines the field of view of said first and second image capture means (1A, 1B). The imaging system is arranged so that the first image capture means (1A) has a field of view including a top view and a first upper side view of an article (5) on said conveying means (6) and the second image capture means (1B) has a field of view including a top view and second upper side view of said article (5), wherein said second upper side view opposes said first upper side view. Multiple lanes arranged side by side may be provided and lighting that provides substantially uniform light over the article (5) is also provided.

Abstract: A method for estimating motion of each of a plurality of tessels in an intermediate image relative to a reference image, which includes searching the reference image to find points that lie along epipolar lines in the reference image corresponding to upper-left and lower-right vertices of the tessel, respectively, that result in a best-matching shape; estimating a depth of each of at least two of the vertices of the tessel; and using the depth estimates of the at least two vertices of the tessel to estimate the motion of the tessel relative to the best-matching shape.

Abstract: The present invention is a method and apparatus to segment a three-dimensional scene having foreground and background regions. Regions in stereo images of the 3-D scene are matched. The stereo images include left and right images. Ranges of matched regions are computed. The foreground region is separated from the background region using the computed ranges of the matched regions.

Abstract: A client-server system and method that enables efficient, low bit rate transmission of image data over a network from an image server (e.g., active cameras) to a client for, e.g., distributed surveillance. A detection and tracking module detects a human presence within an observation area and provides 2-dimensional face coordinates and its estimated scale to a video transmission module. The captured video is then efficiently encoded in log-polar coordinates using an adaptive log-polar mapping, with a foveation point centered on the face. A fovea region, which covers the target object (face), is uniformly sampled and transmitted at full resolution to the client. The periphery region (background) is sampled according to a log-polar grid.

Abstract: A two-dimensional (2-D) machine-vision safety method and apparatus performs high-integrity, high efficiency machine vision. The system digitally filters a configurable number of time-sequenced camera images of a viewed area to create a filtered image. Objects within view in a training phase are detected and compared to objects detected in a runtime phase. Each new image is then compared to the previously created model and a pixel-oriented difference image is calculated. The difference image is then processed for segmentation features attributable to an intruder object. Alarm conditions are determined from threshold parameters related to location of the intruder objects. For segmentation results that exceed a predetermined threshold, an alarm condition is set. Setup parameters are automatically generated by analysis of distinctively marked physical boundaries in an image.

Abstract: Techniques for compressing and decompressing images captured from viewpoints throughout an n-dimensional space. As opposed to conventional approaches, these techniques exploit image coherence over the entire n-dimensional space, providing quick access to images along arbitrary contiguous viewpoint paths during computer graphics applications. This is accomplished by combining a hierarchical data structure with a compression methodology. Such techniques provide for generating and arranging reference images and residual images into a spatial image hierarchy such as a binary tree structure. Images and residuals may be encoded using intra-image compression. A captured image is extracted via a sequence of image warping and addition operations.

Abstract: To provide a variety of methods for extracting digital watermark information from an image as correctly as possible. For example, a first pattern arrangement and a second pattern arrangement which is different from the first pattern arrangement for use in a calculation for extracting the digital watermark information are provided. A state of an original image without the digital watermark information embedded is inferred, using a second pattern arrangement, and a reliability of the digital watermark information to be extracted in accordance with the state is determined. Or to decide the embedding position of the digital watermark information from the image data, the extracting start position is discriminated by detecting the positional information by multiple times at the different extracting start positions. Or to detect a geometrical distortion, a registration signal is embedded on the image.

Abstract: A method and apparatus for decoding an image data comprising decoding a coded map signal indicative of a position and shape of a content in a screen inputted for every block of picture signals, decoding coded orthogonal transform coefficients; performing an inverse orthogonal transform of the orthogonal transform coefficient; and deriving a regenerative picture signal on the basis of the results of the inverse orthogonal transform, wherein a two-dimensional orthogonal transform of all the coefficients with respect to blocks located inside of a content among the orthogonal transform coefficients is performed, and the inverse orthogonal transform of only the coefficients contained in the content with respect to the blocks containing a boundary portion of the content is performed.

Abstract: In a video encoding apparatus 1, the conversion unit 8 decomposes a coding target image by a Matching Pursuits (MP) method using a basis set based on a two-dimensional function which is made by giving parameters for curving a predetermined two-dimensional pattern to a generating function for generating such two-dimensional pattern. According to such configuration, it becomes possible to decompose a coding target image by a small number of repetitive operations on the basis of the MP method. As a result, a coding target image can be decomposed into a small number of atom information, and thereby it becomes possible to generate compression data with a low bit rate.

Abstract: A system and method for facilitating low bandwidth video image transmission in video conferencing systems. A target is acquired (video image of a person's head) and processed to identify one or more sub-regions (e.g., background, eyes, mouth and head). The invention incorporates a fast feature matching methodology to match a current sub-region with previously stored sub-regions. If a match is found, an instruction is sent to the receiving computer to generate the next frame of video data from the previously stored blocks utilizing a texture synthesis technique. The invention is applicable for video conferencing in low bandwidth environments.

Abstract: A system and method analyzes a compressed video including a sequence of frames. The amount of a dominant feature in each frame of the compressed video is measured. A label is associated with each frame according the measured amount of the dominant feature. Views in the video are identified according to the labels, and the video is segmented into actions according to the views. The video can then be analyzed according to the action to determine significant events in the video.

Abstract: A coding process is enabled in object units for an image signal having no shape information, without performing a process for extracting shape information of a specific area in a natural image from the image signal having no shape information corresponding to an image, such as the natural image. The image coding apparatus has a shape value data memory which contains shape values independent of pixel values constituting an image signal having no shape information. The coding of the pixel value is performed on the basis of the shape value read from the shape value data memory, as well as the coding of the shape value being performed, and then pixel coded data and shape coded data are multiplexed and output.

Abstract: A method of compressing a stream of pixel data relative a two-dimensional object, pixels of which are scanned by rows from a source device to a receiver device, includes defining an extended context window to include a pair of pixels following a last encoded pixel on the row being scanned and the respective triplets of neighboring pixels belonging to the preceding row. The method includes defining a first distinct context array of pixels of the extended context window for the pixel of the pair immediately following the last encoded pixel, and a second context array of pixels of the extended context window for the other pixel of the pair. An extended context value relative to each pixel of the pair is calculated, and the extended context value relative to a first pixel immediately following the last encoded pixel is compared with an extended threshold.

Abstract: Various systems, methods, and programs embodied in computer readable mediums are provided for compression of a page frame that includes a number of pixels. In one embodiment, an object map is generated that includes a number of values, each of the values being associated with a corresponding one of the pixels in the page frame. Also, a compressed image file is generated of the page frame that is associated with the object map. Each of the values in the object map indicates whether a blank color conversion or one of a number of predefined color conversions is to be performed during a reconstruction of a corresponding one of the pixels, wherein each of the predefined color conversions includes a corresponding degree of gray component replacement.

Abstract: A specific frame is selected from a plurality of frames of moving image data, and an initial contour corresponding to an extraction target is generated in the selected frame. The initial contour is converged to the contour of the extraction target. An initial contour corresponding to the extraction target is set in an other frame on the basis of the result of the convergence. Contour convergence to the contour of the extraction target is effected on the basis of the set initial contour. As the other frame, frames preceding and subsequent to the selected frame with respect to time can be set.

Abstract: It cannot be ensured that display setting changes are made at once in a system in which the number of display settings for a display processing apparatus is extremely large, an application is running under an operating system and display setting register change time is unpredictable from the application due to another task, or an internal configuration is incapable of immediately reflecting access from a host CPU to a display setting register. According to the present invention, outputs of a plurality of first display setting registers mapped into an address space for temporarily holding display setting data generated and outputted by a host CPU under control of a write control signal are connected to inputs of a plurality of second display setting registers holding display setting parameters referenced by display output module.

Abstract: A shape information coder divides the pixels in the shape information into subsets representing different subsamplings of the shape information. Each pixel in each subset is coded with reference to a context which may be derived from reference pixels in more than one of the subsets. The context preferably includes reference pixels located on all sides of the pixel being coded, which leads to efficient coding. The shape information coder may have two or more cascaded stages, each operating as above, with one subset of pixels passed from each stage to be coded in the next stage. The resulting coded shape information is useful in, for example, wavelet coding of pictures.

Abstract: In a method and a system for encoding and transmission of still images having at least one region of interest (ROI), the ROI coefficients of an image transformed into the frequency domain, preferably using a wavelet transform are encoded so that they are transmitted first and can be decoded by a receiver without transmission of the boundary of the ROI. In a preferred embodiment the coefficients belonging to the ROI are shifted so that the minimum ROI coefficient is larger than the largest background coefficient. A receiver can then perform an opposite procedure and thereby obtain the ROI. By specifying how much the coefficients need to be shifted in order to avoid sending shape information several advantages are achieved. Thus, it is possible to avoid sending shape information and to avoid shape encoding at encoder side. Furthermore, there is no need for a shape decoder at receiver side, and there is no need for the receiver to produce the ROI mask.

Abstract: An image processing apparatus and method for deriving a digital watermark embedded in an image is provided which has a recognizing unit for recognizing a predetermined characteristic area in the image and a deriving unit for extracting the digital watermark from the image data in the recognized characteristic area.

Abstract: A video system includes a plurality of frames of video each of which is defined by a plurality of scene elements. The scene elements for a respective frame together define an image of the frame. First auxiliary data is descriptive of a first scene element of the frame and second auxiliary data is descriptive of a second scene element of the frame. A sending device sends the frame of video, including its scene elements, the first auxiliary data, and the second auxiliary data, from the sending device to a receiving device.

Abstract: This invention provides an image processing apparatus/method characterized by inputting successive image data, dividing the image data into blocks each formed from a plurality of pixels, detecting the motion vector of each block, calculating a probability at which the block has a predetermined area label, in accordance with the motion vector and the area label of a neighboring block, determining the area label of each block in accordance with the calculation result, and dividing the image data into areas of respective objects in accordance with the determined area label.

Abstract: A method of encoding images (1) including areas (11) of relevant pixels and areas (12) of irrelevant pixels (e.g. a background) is disclosed. The pixels of the irrelevant area are replaced (4) by pseudo-pixels smoothing the transition between said areas. The image thus obtained is subsequently subjected to conventional image coding (5) which now yields considerably fewer bits. At the receiver end, the pseudo-pixels are replaced (9) by predetermined background pixels to reconstruct the original boundaries (13). If the boundary between both areas is not known in advance, the boundary is detected (2), encoded (3) and transmitted to the receiver. The invention is particularly applicable to compression of medical images having circular areas of interest, but may also be used to encode other video images having irrelevant areas such as videophone images in which a head-and-shoulder area constitutes relevant information only.

Abstract: An image data compression technique includes division of an image data stream into a plurality of subregions and optimal compression of each subregion. The subregions may represent adjacent pixels in a reconstructed image, and may all be of the same length. The length of the subregions may be set to a default, and may be altered based upon image characteristics. The subregions are analyzed for selection of an optimal compression algorithm for each subregion from a set of candidate algorithms. Criteria for selection of the optimal algorithms may include the relative entropy levels of data within the subregions, and the length of compressed data code resulting from application of each of the candidate algorithms.

Abstract: A system and method for compressing a video image. The system includes an object database and a compression processor coupled to the object database for receiving video image information from a video source. The compression processor includes an image analyzer for determining status of one or more pixels of the video image over time, a shape determiner for determining a shape that defines a group of pixels that are determined to have the same status over a period of time, an object retriever for retrieving an object from the object database that corresponds to the determined shape from a stored set of objects, and a saving component for saving information relating to the selected object and dimensional information of the determined shape.

Abstract: The present invention provides a transmission apparatus for video information that allows wireless transmission of video information at a low rate regardless of differences in OSs.

Abstract: In a method and a device for transmission of S+P transform coded digitized images a mask is calculated by means of which a region of interest (ROI) can be transmitted lossless whereby the ROI can be transmitted and received lossless and still maintaining a good compression ratio for the image as a whole. This is possible since no or very few bits can be used for the remaining part of the image. The calculated mask can also be used for transmitting the coefficients needed for a lossless region of interest during any stage of the transmission.

Abstract: This invention relates to encoding and decoding techniques concerning images. The invention utilizes the principle that an image or a part thereof may contain a dominating direction that is other than a horizontal or vertical direction Using DCT components, which are designed to this dominating spatial frequency, fewer non-zero components are needed for the transformation than if the normal DCT components are used, thereby enabling more efficient compression of the image.

Abstract: A method extracts high-level features from a video including a sequence of frames. Low-level features are extracted from each frame of the video. Each frame of the video is labeled according to the extracted low-level features to generate sequences of labels. Each sequence of labels is associated with one of the extracted low-level feature. The sequences of labels are analyzed using learning machine learning techniques to extract high-level features of the video.

Abstract: The image coding apparatus is constituted comprising an atom extraction section 109 for acquiring position data indicating the positions of atoms of prediction residual data in macroblocks, for each of the macroblocks; and a variable length encoding section 113 for determining the frequency of occurrence distribution of the position data in the macroblocks to be encoded, in accordance with the number of atoms of prediction residual data present in the macroblocks, and for performing arithmetic coding of the position data on the basis of the frequency of occurrence distribution thus determined. An improvement in the efficiency of the entropy coding of the atom parameters of Matching Pursuits coding is therefore feasible.

Abstract: A decoder decodes encoded binary image data on each block in order to recover a target block of the binary image. The image decoder includes motion compensated blocking which obtains a reference block from a reference binary image by applying motion compensation using motion information. A selector selects a statistical model from among a plurality of statistical models, based on states of pixels surrounding a reference pixel in the reference block, in which the reference pixel also corresponds to a target pixel in the target block. An arithmetic decoder recovers the target block by decoding the encoded data using the selected statistical model. In one embodiment, the pixels surrounding the reference pixel in the reference block are pixels positioned within one pixel distance from the reference pixel.

Abstract: A method of transmitting selected regions of interest of digital video data at selected resolutions, including: capturing digital video; compressing the digital video into a sequence of individual high resolution JPEG 2000 frames and simultaneously extracting from the compressed high resolution JPEG 2000 frames a lower resolution; storing the individual high resolution JPEG 2000 frames in a storage device; creating a video sequence of the lower resolution representation; transmitting the video sequence of the lower resolution representation to a user; creating one or more video sequences of selected regions of interest at selected resolutions greater than the resolution of the video sequence of the lower resolution representation; transmitting one or more video sequences of selected regions of interest at selected resolutions; and repeating the prior two steps at incrementally increased resolutions until a desired resolution of the selected region of interest is reached according to a viewing objective.

Abstract: There is provided an image decoding apparatus for decoding a coded image signal having an image identifier based on a data structure of the image signal. This apparatus comprises a data analyzer for analyzing the coded image signal with reference to the image identifier to decide whether the coded image signal is a coded arbitrary shape signal including both of coded shape bit streams obtained by coding a shape signal which represents the shape of each object as one of the components of a display image and coded pixel value bit streams obtained by coding a pixel value signal representing the gradation of the object, or a coded binary signal including only coded shape bit streams obtained by coding a shape signal representing a display image of binary information; a first decoder for decoding the coded shape bit streams by a first decoding process; and a second decoder for decoding the coded pixel value bit streams by a second decoding process.

Abstract: A method and system for organizing and storing map data that facilitates use of the map data by navigation application programs such as those in navigation systems. The map data represent geographic features located in a region. The map data are organized into parcels such that each parcel contains a portion of the map data. The map data contained in each parcel represent those geographic features contained in a corresponding separate one of a plurality of rectangular areas located in the region. Each rectangular area has a uniform dimension in a first coordinate direction but has a dimension in the other coordinate direction such that the map data that represent the geographic features contained in the rectangular area are close to, but do not exceed, a maximum data size for a parcel.

Abstract: A moving picture coding method capable of maintaining high picture quality for an important area even in a low bit rate and gradually improving picture quality of the neighboring area as the bit rate becomes higher. According to this method, the important area detection section 122 automatically detects an important area within a frame, the gradual shift map generation section 124 generates a gradual shift map whose shift value decreases gradually from the important area toward the neighboring area. The bit shift section 130 bit-shifts DCT coefficients according to the gradual shift map. In this way, more DCT coefficients which contribute to improvement of picture quality of the important area are stored in the start portion of the enhancement layer preferentially.

Abstract: An image of a compound document is processed by accessing a plurality of blocks of the image; and classifying the blocks for compression. At least some of the blocks are classified according to their numbers of distinct colors.

Abstract: In a method of processing video into an encoded bitstream in which the encoded bitstream is intended to be sent over a WAN to a device, the processing of the video results in the bitstream (a) representing the video in a vector graphic format with quality labels which are device independent, and also (b) being decodable at the device to display, at a quality determined by the resource constraints of the device, a vector graphics based representation of the video.

Abstract: There are provided an improved scan interleaving method based on context values of neighboring pixels, an improved binary shape encoding/decoding method, and encoding/decoding apparatus suitable for the same. The hierarchical binary shape encoding apparatus includes a portion for down-sampling for SI method, the portion for performing down-sampling for SI method on the binary shape data to form a base layer, a portion for encoding the base layer supplied from the down-sampling portion, and a portion for encoding an enhancement layer having the current binary shape data by referring to the base layer or a preceding frame. The encoding apparatus can perform hierarchical encoding, thereby enhancing the overall coding efficiency.

Abstract: In a high-frequency component processing section, a low-quality color transformation section applies a very simple (meaning a small amount of calculations) color transformation to an original image as required and a high-frequency component calculation section then applies out frequency transformation thereto. In a low-frequency component processing section, an LPF (Low-Pass Filter) and a sampling process executed by a sub-sampling section cooperate in sub-sampling the original image to obtain a reduced image with a small number of pixels. A high-quality color transformation section applies an advanced (meaning a large amount of calculations) color transformation to the reduced image. Then, a low-frequency component calculation section applies frequency transformation to the reduced image, which has been subjected to the color transformation.

Abstract: Video images in, e.g., a multimedia stream are scanned prior to compression for transmission to recognize symbols, such as graphics symbols and alpha-numeric characters. The types, positions, sizes, etc. of the symbols are recorded to render symbol information, and then the images are compressed with or without compressing the symbols, which may be removed from the images prior to compression if desired. The compressed video and symbol information are sent to a receiver, which decompresses the video, optionally transforms the symbols, and then inserts the symbols where indicated by the symbol information.

Abstract: There is provided an image decoding apparatus for decoding a coded image signal having an image identifier based on a data structure of the image signal. This apparatus comprises a data analyzer for analyzing the coded image signal with reference to the image identifier to decide whether the coded image signal is a coded arbitrary shape signal including both of coded shape bit streams obtained by coding a shape signal which represents the shape of each object as one of the components of a display image and coded pixel value bit streams obtained by coding a pixel value signal representing the gradation of the object, or a coded binary signal including only coded shape bit streams obtained by coding a shape signal representing a display image of binary information; a first decoder for decoding the coded shape bit streams by a first decoding process; and a second decoder for decoding the coded pixel value bit streams by a second decoding process.

Abstract: A gray scale image or a color image is entered for wavelet transformation. A layout feature is determined from a coefficient of the lowest frequency band after the wavelet transformation. A shape feature is determined from a coefficient of the second lowest frequency band after wavelet transformation. Furthermore, a texture feature is determined from the average and deviation of frequency bands excepting the lowest frequency band after wavelet transformation.

Abstract: It is known to represent a square block of an image which has a maximum horizontal or vertical extension N as N vectors each representing one column or row, and encode it using an N×N discrete cosine transform matrix. We introduce a technique which transforms an input vector, or array, representing a portion of an arbitrarily-shaped object block in which the maximum horizontal or vertical extension is at most N, by adding to it additional padding elements, to produce a vector or array which can be encoded using an N×N discrete cosine transform. The padding maximises the number of transformed elements which are zero. We propose encoding arbitrarily-shaped objects based on this padding technique.

Abstract: System consistent with this invention comprise a first coding part configured to code a map signal indicative of the position and shape of a content in a screen inputted for every square block of picture signals to be coded, a subband dividing part configured to divide the picture signals into a plurality of subband picture signals, a resolution transform part configured to perform the resolution transform of the map signal into the resolution of each of the subband picture signals divided by the subband dividing part, and a second coding part configured to code each of the subband picture signals divided by the subband dividing part when blocks are inside of the content, and coding only the signals inside of the content when the blocks contain the boundary portion of the content in the subband picture signals according to the map signal resolution-transformed by the resolution transform part.

Abstract: A moving picture coder has a first coding mode in which the entire picture is coded and a second coding mode in which a partial region of the picture is coded. Coding is so performed that the frame rate inside the partial region is higher than that outside the partial region. To perform such coding, the coder comprises a region selecting section (201) for selecting the partial region, a shape coding section (202) for coding is the shape of the partial region, a moving picture coding section (204) for coding only the values of the pixels inside the partial region according to the selected region, and a coding control section (205). A moving picture decoder decodes a picture in the second mode by superimposing decoded picture data of the partial region of the picture in the second mode on the decoded picture data of the entire picture in the first mode.