Abstract: The invention provides a method, data compression system, apparatus, and article of manufacture which reduce the error in transform equations in which constants are replaced by approximations. According to the invention transform constants are replaced with approximations which are a function of an integer and a floating point value. The transform equation is then performed with the integers in place of the constants. The floating point value may be applied either to the result of the equation or to the data to be processed by the transform equation before the equation is performed. Further the floating point value may be applied using a fixed precision version of the value or a fixed precision value into which the floating point value has been factored.

Abstract: An image forming apparatus, which includes a plurality of image forming portions transferring a yellow image, a magenta image, a cyan image, and a black image formed on a plurality of photoconductor drums to a sheet conveyed on a conveying belt, a marking unit forming marks on the conveying belt, a detecting unit detecting the marks with three or more sensors aligned in a direction normal to a direction in which the sheet is conveyed, a calculating unit calculating an amount of color misalignment in accordance with results detected by the detecting unit, and a correcting unit correcting the color misalignment in accordance with the calculated amount of color misalignment, wherein the calculating unit calculates an amount of skew difference in accordance with results detected by two sensors among the three or more sensors, wherein one sensor of the two sensors is disposed on one end of the three or more sensors and the other sensor of the two sensors is disposed on the other end of the three or more sensors, w

Abstract: In a parallel computation of a Hough transform of an array of input data values, the transform space of the Hough transform is partitioned dynamically or statically into a number of sub-spaces. Each sub-space of the transform is stored in a sub-space of memory locations. Data values from the array of input data values are passed to a plurality of processors, each processor associated dynamically or statically with a sub-space of memory locations. Each processor, acting in parallel with the other processors, updates constituent elements of the Hough transform stored in the associated sub-space memory locations dependent upon the input data value.

Abstract: The present invention provides a land mark detecting apparatus simultaneously used with a land mark comprising plural mark primitives. The land mark detecting apparatus comprises a set of cameras and a land mark detector which includes an image signal interface unit, a pre-processing unit, a straight line detector to and a mark specifying unit.

Abstract: Computer software for and a method of determining location of a target image within a two-dimensional input image by employing a three-dimensional reference image comprising determining object edge points in the input image; given estimated seeker line-of-sight orientation and range data, computing a two-dimensional reference image from the three-dimensional reference image; applying a weighted-direction-cross-product Hough Transform to the object edge points and points of the two-dimensional reference image; classifying the input image as containing a target image or not; if the image contains a target image, identifying a location of the target image within the two-dimensional input image; computing confidence measure as to the selected location of the target image; and in the case of sequential decision, also performing confidence accumulation through multi-frames or multi-looks.

Abstract: A method and apparatus are provided to measure spatial distortion. Measure and remove borders; and rescale the active test image, the active reference image, or both, such that they are the same size. A representative line is obtained, or selected, from each image providing a reference representative line (ReferenceRL) and a test representative line (TestRL). A local Pearson' cross-correlation coefficient (LPCCC) image is created from both the ReferenceRL and the TestRL at different horizontal shifts. The highest average intensity line across the LPCCC image is found as the coordinates of the corresponding Hough Transform image pixel with maximum intensity is determined. The Hough Transform Theta and R are converted to spatial scale and offset. The crude and refined offsets and scale values are combined to produce total horizontal and vertical scale offset values. These spatial measurements can then be provided.

Abstract: Datawords encoded using a first-order Reed-Muller code are decoded using a Fast Hadamard Transform (FHT). Where the dataword is known to comprise a number of unused bits at one end, the soft decisions of the codeword are reordered. Majority polling is then performed on groups of the soft decisions to produce a series of soft decisions which make up a new codeword. The new codeword is then decoded using an FHT of lower order than the FHT applicable to the originally received codeword.

Abstract: In some embodiments, a method for performing and a system configured to perform a 2D transform (for example, an inverse discrete cosine transform) on each block of a sequence of data blocks, where the 2D transform includes a row transform and a column transform. To perform the row or column transform on a row or column of data, these embodiments determine whether each of different subsets of the data values comprising a partition of the row (column) includes at least one zero value, whether each of different subsets of a first subset of the partition includes at least one zero value, and whether each of different subsets of at least one other subset of the partition includes at least one zero value.

Abstract: The camera includes a sensor for sensing the photographer's iris image and registering the image in advance. The iris image is recorded in the image of a subject by a digital MCU at a timing different from that at which the image of the subject is captured. The recording timing is that at which the camera power supply is turned off, that at which a recording medium is ejected from the camera or that at which the iris image to be recorded is changed to the registered iris image of another photographer. The recording of the iris image is achieved by embedding it as a watermark or by appending it to metadata.

Abstract: A method is described that adjusts the position of a item and sets a tilt angle for each of a plurality of micro-mirrors of a digital micro-mirror device. The setting of the tilt angles is to establish a filter within the optical channel of an inspection tool that inspects the item. The filter is to reduce noise received at an optical detection device. The tilt angle settings are a function of the position. The method also includes comparing information from the optical detection device that describes an inspected region of the item's surface against an expected version of the information.

Abstract: An implementation of Hough transform is described. According to one aspect, the Hough transform is used for line detection in an image based on a one-dimensional voting array.

Abstract: An implementation of Hough transform is described. According to one aspect, the Hough transform is used to perform motion analysis of patterns in a video using one-dimensional detectors and a one-dimensional voting array.

Abstract: A process for color graphics image processing, related to detection and segmentation of sweeps, is provided. An input graphics image is transformed into a three-dimensional histogram in an appropriate color space 104 (e.g., CIELUV). Two-dimensional histograms are estimated from the three-dimensional histogram 106. The two-dimensional histograms are processed to detect and segment sweeps 108. Sweep segment information from the processing of the two-dimensional histograms is combined 110. The combined sweep segment information is used to process the input graphics image to identify and segment sweeps 112. Post-processing may be optionally and selectively used to reject false alarms (i.e., areas falsely identified as sweeps) 114.

Abstract: In order to segment a digital image of a postal packet in order to read a postal address (AP) automatically on said packet, the method consists: in subdividing (20) the image into image blocks of identical size; in applying (30) processing based on a Hough transform to each image block in order to identify in the image block in question a privileged direction for the distribution of image points in the block under consideration; in grouping together (40) contiguous image blocks within the image having a common privileged direction in order to define (50) a rectangular zone of interest (ZI) containing the group of said contiguous blocks having a common privileged direction in such a manner that said rectangular zone of interest has a longitudinal edge parallel to said common privileged direction; and in applying the automatic address-recognition algorithm to said zone of interest in the image in order to read a postal address automatically.

Abstract: An implementation of Hough transform is described. According to one aspect, the Hough transform is used to perform motion analysis of patterns in a video using two-dimensional primary data items and a two-dimensional voting array.

Abstract: A method identifies edges of an object in an image by analyzing an accumulator array produced via a Hough transformation of the image. The method includes the step of identifying a first cell in a first row of the array associated with a first edge of the object. The first is then searched for a second cell in the array associated with a second edge of the object, the second edge being parallel to the first edge. A second row shifted ninety degrees from the first row is then searched for a third cell in the array associated with a third edge of the object, the third edge being perpendicular to the first edge. The second row is then searched for a fourth cell in the array associated with a fourth edge in the image, the fourth edge being parallel to the third edge.

Abstract: The present invention provides an image processor and an image processing method wherein a specific image such as an image of which the printing is prohibited can be recognized at high speed and in addition, with high precision. A rendering command analysis unit passes a rendering object to any of a recognition unit for a raster image, a recognition unit for graphics and a recognition unit for a character according to the type of the rendering object in the rendering command and tries to recognize a specific image. The result of recognition by each recognition unit is passed to a recognition determination unit and it is synthetically determined based upon these results of recognition whether the specific image exists or not. Even if the specific image is formed by dividing into plural types of rendering objects, the specific image can be recognized by such recognition processing.

Abstract: A system and method for tracking multiple targets using distributed linear sensor arrays. The system includes a plurality of arrays of sensors for receiving signals from a target; a receiver for receiving signals received by the plurality of sensor arrays; an analog/digital converter for converting the signals received from the sensor arrays to a digital format, if signals are received in an analog format; a digital storage device for storing the digitized data from the sensor arrays; and a computer system for retrieving the stored digitized data from the plurality of sensor arrays and processing the data through the use of a composite Hough transform to determine a track of the target. Embodiments include a computer system and method for processing target data from sensors through the use of a composite Hough transform to determine a track of the target.

Abstract: An Hadamard transform coding circuit changes a rounding method for each Hadamard transform block to prevent a rounding operation from providing an error biased in one direction, as seen in a picture in its entirety, for data compression and rounding. More specifically, after Hadamard transform there exist decimal parts 0.0, 0.25, 0.5 and 0.75, and if a numerical value is simply rounded off a probability would be increased that the numerical value is biased to increase in absolute value. Accordingly for 0.5 a rounding operation to provide an integer switches for each block. Thus a motion image decoding apparatus can be provided to reduce error accumulation and suppress flickering, color variation and other similar noticeable degradations.

Abstract: A Hough transform unit executes Hough transform to HIGH pixels of outline binary image data inputted thereto, and stores the calculation result in a Hough space data storage. A Hough space data calculating/projecting unit sequentially reads out data stored in the Hough space data storage, executes a specific calculation, and thereafter stores the calculation result sequentially in a calculated projection data storage. An angle detector sequentially reads out calculated frequency data stored in the calculated projection data storage, calculates the maximal value of the data read out, and detects an angle that gives the maximal value as the skew angle. The image processing device, being thus configured, allows detecting and correcting the skew angle with high accuracy, even when the input image contains image elements such as photograph images and dot images.

Abstract: A staring imaging system inputs a 2D spatial image containing multi-frequency spectral information. This image is encoded in one dimension of the image with a cyclic Hadamarid S-matrix. The resulting image is detecting with a spatial 2D detector; and a computer applies a Hadamard transform to recover the encoded image.

Abstract: An analyzer/classifier tool (100) for data comprises use of an energy minimization process (120) with data transformed by an input process (110) into one or more input matrices. The data to be analyzed/classified is processed by an energy minimization technique such as individual differences multidimensional scaling (IDMDS) to produce at least a rate of change of stress/energy. Using the rate of change of stress/energy and possibly other IDMDS output, a back end process (130) analyzes and classifies data through patterns recognized within the data.

Abstract: A system for efficient bit plane coding of transform coefficient data, such as DCT data used in a video coding system. Decimal values for the transform coefficients are converted to binary values, where each bit occupies a corresponding bit plane, from the most significant bit to the least significant bit. One bit from each coefficient is provided in a common bit plane. A one-bit flag or codeword is used for coding one or more initial all-zero bit planes, while another one-bit flag is used for designating the first subsequent non-all-zero plane. For the first non-all-zero plane, a reduced coding table is used to provide codewords that follow the one-bit flag.

Abstract: Backscatter imaging using Hadamard transform masking includes an area x-ray source with alternating, masked, Hadamard transform patterns. The total backscatter signal from a target for each pair of corresponding masks is recorded. The difference in signal strengths for each pair of corresponding masks is a direct measurement of the Hadamard transform coefficient for that mask. An image of the target is formed by performing an inverse discrete Hadamard transform on the complete matrix of coefficients.

Abstract: A method for locating symbols arranged in one or more rows in an image includes smearing the image, and fitting line segments through edge points of features in the smeared image. A group is found of the line segments that are in mutual proximity and are mutually substantially parallel. A region of the image that contains the group of the line segments is identified as a possible location of the symbols.

Abstract: A method and system for the extraction of linear features from digital imagery is described herein. The proposed methodology is based on the tridimensional Radon transformation and on the generation of a multi-layer database having separate layers containing different types of information to allow the reconstruction of an output image where the linear features of the input image are restored.

Abstract: A method of document segmentation. Specifically, one embodiment of the present invention discloses a method of document segmentation that performs a plurality of projection profiles of pixel intensities on a document containing a plurality of text lines over a range of angles. A plurality of slope values for a plurality of discrete distances perpendicular to said range of angles is calculated for the plurality of projection profiles. A set of maximum absolute slope values is sorted out from the plurality of slope values. Text lines of first and second type are identified by setting a threshold slope value. Absolute slope values greater than the threshold slope value indicate the plurality of text lines of said first type. Absolute slope values less than the threshold slope value indicate the plurality of text lines of a second type.

Abstract: A wavelet domain half-pixel motion compensation process that reduces aliasing effects that down sampling causes in the wavelet transform uses an H-transform and provides motion estimation and compensation in wavelet domain without requiring an inverse wavelet transform. For encoding, a q-dimensional (e.g., q=2) H-transform is applied in a conventional manner to non-overlapping q×q matrices in a first frame. When determining motion vectors for a second frame, “half-pixel” interpolation of the wavelet data of the first frame determines generates half-pixel data corresponding to q×q space-domain matrices that are offset (e.g., 1 pixel) horizontally and/or vertically from the q×q matrices that were transformed. Motion estimation techniques can then identify object motion by comparing wavelet domain object data in one frame to actual and interpolated wavelet domain data for another frame.

Abstract: A display of code power levels and relationships of multiple spreading factor orthogonal codes in a code division multiple access (CDMA) signal has a first code graph of code domain power (CDP) for all the codes at the largest code spreading factor that includes all codes at all spreading factors plotted in a linear code indexing order and a second graph of CDP where the codes are plotted in bit-reversed order to group together the sub-codes of each code of lower spreading factor. Markers on each graph are tied together such that movement of one marker on one graph to a particular code value causes the other marker on the other graph to move to the corresponding code value. The sub-codes related to a particular code bar in the graphs indicated by the markers may be differentiated by using grey scale or color variations.

Abstract: A system and process for recognizing an object in an input image involving first generating training images depicting the object. A set of prototype edge features is created that collectively represent the edge pixel patterns encountered within a sub-window centered on each pixel depicting an edge of the object in the training images. Next, a Hough kernel is defined for each prototype edge feature in the form of a set of offset vectors representing the distance and direction, from each edge pixel having an associated sub-window exhibiting an edge pixel pattern best represented by the prototype edge feature, to a prescribed reference point on a surface of the object. The offset vectors are represented as originating at a central point of the kernel. For each edge pixel in the input image, the prototype edge feature which best represents the edge pixel pattern exhibited within the sub-window centered on the edge pixel is identified.

Abstract: The Hough Transform is a computer vision algorithm that can robustly detect a wide variety of features such as lines, circles, and anything else that can be readily parameterized or otherwise cast in terms of a discrete popularity algorithm. Unfortunately, not all processors or like devices are capable of providing the requisite processing capability usually associated with the Hough Transform. The methods and arrangements presented herein leverage the dedicated hardware of a graphics card to provide a portion of the data processing.

Abstract: A method for automatic association of moving target indications from at least one entity traveling along a route. A moving target indicator radar is used to detect a plurality of moving target indication data. The moving target indication data proximate to the identified route is selected and presented in a distance-time graph, such that each selected moving target indication data has a unique distance along route and a unique observation-time value. The selected moving target indication data are then transformed from the distance-time coordinate to a slope-intercept coordinate, such that co-linear moving target indication data in the distance-time coordinate are transformed into a plurality of points superposed together with nearly identical slope values and nearly identical distance intercept value. The superposed points are mapped back to the distance-time coordinate, and the moving target indication data corresponding to the superposed points are thus associated.

Abstract: A system and process for recognizing an object in an input image involving first generating training images depicting the object. A set of prototype edge features is created that collectively represent the edge pixel patterns encountered within a sub-window centered on each pixel depicting an edge of the object in the training images. Next, a Hough kernel is defined for each prototype edge feature in the form of a set of offset vectors representing the distance and direction, from each edge pixel having an associated sub-window exhibiting an edge pixel pattern best represented by the prototype edge feature, to a prescribed reference point on a surface of the object. The offset vectors are represented as originating at a central point of the kernel. For each edge pixel in the input image, the prototype edge feature which best represents the edge pixel pattern exhibited within the sub-window centered on the edge pixel is identified.

Abstract: A coding apparatus which can simplify a coding circuit and also simplify the processing of quantization. Image data that consists of R-, G-, and B-color components read from a scanner into buffers is subjected to transformation using a position correlation for each of RGB in position-correlation transformation sections. The DC component of each of the components is then color-converted in a color conversion section to be a DC lightness signal and DC color signals, and the signals are quantized in a quantization section and a coding section. Whereas, the AC components of each of the components are subjected to quantization in a first step in an AC component quantization section, and further to color conversion in a color conversion section to be an AC lightness signal and AC color signals. The signals are quantized in a second step in a quantization section and a coding section.

Abstract: The image processing device includes: a distance image capture device which captures a distance image based upon three dimensional coordinate values obtained, by measuring the distances to various points upon body surfaces in the forward visual field of a distance sensor, for those points upon those body surfaces; a straight line extraction device which obtains three dimensional coordinate values of two points from a viewpoint picture element in the distance image and a picture element which is separated by a predetermined distance from that viewpoint picture element, and which extracts a plurality of straight lines in a horizontal plane using two dimensional coordinate values by ignoring the vertical component among these three dimensional coordinate values; and a plane detection device which obtains points on the body surfaces which lie upon various straight lines which have been extracted by the straight line extraction device, and detects points which lie upon the same straight line as being upon the same

Abstract: A method of analysing an image comprises performing a Hough transform on points in an image space to an n-dimensional Hough space, selecting points in the Hough space representing features in the image space, and analysing m of the n variables for the selected points, where m is less than n, for information about the features in the image space.

Abstract: The present invention comprises using error propagation for building feature spaces with variable uncertainty and using variable-bandwidth mean shift for the analysis of such spaces, to provide peak detection and space partitioning. The invention applies these techniques to construct and analyze Hough spaces for line and geometrical shape detection, as well as to detect objects that are represented by peaks in the Hough space. This invention can be further used for background modeling by taking into account the uncertainty of the transformed image color and uncertainty of the motion flow. Furthermore, the invention can be used to segment video data in invariant spaces, by propagating the uncertainty from the original space and using the variable-bandwidth mean shift to detect peaks. The invention can be used in a variety of applications such as medical, surveillance, monitoring, automotive, augmented reality, and inspection.

Abstract: Datawords encoded using a first-order Reed-Muller code are decoded using a Fast Hadamard Transform (FHT). Where the dataword is known to comprise a number of unused bits at one end, the soft decisions of the codeword are reordered. Majority polling is then performed on groups of the soft decisions to produce a series of soft decisions which make up a new codeword. The new codeword is then decoded using an FHT of lower order than the FHT applicable to the originally received codeword.

Abstract: A method for transmitting a vector having at least two vector components, each of the vector components described in a frequency. Each vector is represented as a bit number with a predetermined number of bit levels. The bit numbers are encoded according to a priority of the bit levels and the encoded bit numbers are transmitted.

Abstract: Computer software for and a method of calculating ring-wedge data from a digital image by performing a discrete Fourier transform of the digital image. A discrete autocorrelation, discrete cosine transform, and/or Hadamard transform is also preferably performed, together with providing the results to a neural network (most preferably a fully connected, three-layer, feed-forward neural network with sigmoidal activation functions) to perform pattern recognition on the data.

Abstract: An edge image acquisition apparatus obtains an image of a time (t−&Dgr;t) and an image of time t, i.e., images different in time by &Dgr;t. The apparatus calculates their motion difference to produce a motion difference image. Furthermore the apparatus calculates a spatial difference of one of the images to produce a spatial difference image. The produced motion and spatial difference images are then logically ANDed together to produce a logically ANDed image to extract and output an edge.

Abstract: In order to segment a digital image of a postal packet in order to read a postal address (AP) automatically on said packet, the method consists: in subdividing (20) the image into image blocks of identical size; in applying (30) processing based on a Hough transform to each image block in order to identify in the image block in question a privileged direction for the distribution of image points in the block under consideration; in grouping together (40) contiguous image blocks within the image having a common privileged direction in order to define (50) a rectangular zone of interest (ZI) containing the group of said contiguous blocks having a common privileged direction in such a manner that said rectangular zone of interest has a longitudinal edge parallel to said common privileged direction; and in applying the automatic address-recognition algorithm to said zone of interest in the image in order to read a postal address automatically.

Abstract: An image data transceiver system includes an encoder for converting original image data into processed image data and for transmitting the processed image data. The encoder includes a forward transformation block, having a Modified Hadamard Transform basis for compressing the original image data and for converting original image data into transformed image data. The encoder also includes a progressive transmission block for progressively transmitting the processed image data. The image transceiver system includes a decoder for reconstructing the original image data from the processed image data transmitted by the encoder. A progressive reconstruction block in the decoder progressively reconstructs the transformed image data from processed image data; and an inverse transformation block in the decoder has an inverse Modified Hadamard Transform basis for extracting the original image data from the transformed image data.

Abstract: Apparatus and methods are provided for encoding an input frame of a video sequence for transmission over a channel. The method and apparatus decompose the input frame into multiple subbands and divide the multiple subbands into multiple blocks corresponding to a region of the input frame. The blocks in the highest frequency subbands of the multiple blocks are selected based upon a luminance component of the input frame and the multiple blocks in the highest frequency subbands are classified into a multiple classes to provide a multiple class labels. The multiple class labels are collected to form a subband class map for each of the multiple blocks in the highest frequency subbands and a global class map is constructed from a majority evaluation of the subband map for each of the multiple blocks. The multiple blocks within the multiple subbands are grouped which have one of the class labels to form multiple subband class sequences.

Abstract: A high definition radar imaging system receives SAR image data and adaptively processes the image the data to provide a high resolution SAR image. The imaging technique employs an adaptive filter whose tap weights are computed based upon a constrained Maximum Likelihood Method (MLM). The MLM technique chooses the filter tap weights (i.e., a weighting vector &ohgr;) to satisfy several criteria including: 1) they preserve unity gain for a point scatter at the desired location, and 2) they minimize the perceived energy in the output image. The weights &ohgr; are constrained in norm ∥&ohgr;∥≦&bgr;, to reduce the loss of sensitivity to bright scatters. Significantly, the present invention applies an additional constraint on the iterative selection of the weighting vector &ohgr;, such that the weighting vector &ohgr; shall confined to a particular subspace in order to preserve background information in the image.

Abstract: An implementation of Hough transform is described. According to one aspect, the Hough transform is used to perform motion analysis of patterns in a video using two-dimensional primary data items and a two-dimensional voting array.

Abstract: An implementation of Hough transform is described. According to one aspect, the Hough transform is used to perform motion analysis of patterns in a video using one-dimensional detectors and a one-dimensional voting array.

Abstract: An implementation of Hough transform is described. According to one aspect, the Hough transform is used for line detection in an image based on a one-dimensional voting array.

Abstract: An apparatus and method for correcting the sharpness of an image signal, using a Haar Wavelet transform and a difference between pixel values adjacent to the edge of the image signal to be corrected while reducing the occurrence of overshoot and undershoot at the edge of the image signal, are provided. The apparatus includes an edge detector for detecting data on the edge of the image signal, by performing a multi-stage Haar Wavelet transform on the image signal, a gain detector for detecting a gain for correcting the image edge, a pixel value detector for detecting a corrected pixel value regarding the edge data at a position to be corrected by performing an operation on the edge data, at least one pixel adjacent to the edge data, and the gain, and an image signal generator for generating an image whose edge is formed based on the corrected pixels.

Abstract: A motion compensation coding apparatus using a wavelet transformation and a method thereof are capable of detecting a motion vector with respect to a block having a certain change or a motion in an image from a region having a hierarchical structure based on each frequency band and each sub-frequency band generated by wavelet-transforming an inputted motion picture and effectively coding a motion using the detected motion vector. The motion compensation coding apparatus can include a wavelet transformation unit receiving a video signal and wavelet transforming by regions of different frequency bands based on a hierarchical structure, and a motion compensation unit receiving the wavelet-transformed images and compensating the regions having a certain change or motion in the image.