Abstract: Natural images which are similar to each other contained in a page represented by page description data are corrected so that they have natural appearance to the eye. To achieve this object, an image recognizing unit recognizes images in a page represented by page description data, and a natural image determining unit determines whether or not each recognized image is a natural image. An image analyzing unit calculates a setup condition for image correction for each natural image. A second correction condition calculating unit calculates, for the similar natural images being similar to each other, a correction condition for making image qualities of the similar natural images substantially uniform. An image correcting unit applies image correction based on the setup condition and the correction condition to the similar natural images.

Abstract: A method for generating a non-graphical digital image from an original digital image includes the following steps: performing edge detection to generate a third digital image, performing screen dot detection to detect the photo regions in the original digital image, detecting color regions in the original digital image, removing the photo regions and the color regions from the third digital image to generate the non-graphical digital image.

Abstract: There is described an analysis method for at least one image data record of an examination object, wherein each image data record features a multiplicity of image data elements. A position in a multidimensional space is assigned to each image data element. Each image data element features an image data value. The image data values of positionally corresponding image data elements of the image data records are specified by means of at least essentially positionally identical regions of the examination object. A computer automatically divides the image data records into empty regions and signal regions, applying an overall assignment rule which is based on the image data values of the image data elements of a plurality of image data records, such that each image data element of each image data record is assigned to either its empty region or its signal region.

Abstract: A method and apparatus for partitioning an object from an image, such that substantially the entire object is contained in the partitioned region, includes pre-setting a reference width for the object; extracting a shadow underneath the object from the image and determining a candidate region containing the object in the image based on the extracted shadow underneath an object in the image; acquiring an imaging width of the reference width at the location of the extracted shadow underneath the object in the image, and adjusting the candidate region based on the imaging width of the reference width such that the adjusted candidate region substantially completely contains the object; and partitioning the adjusted candidate region as a region containing the object from the image.

Abstract: A system and method for labeling feature clusters in frames of image data for optical navigation uses distances between feature clusters in a current frame of image data and feature clusters in a previous frame of image data to label the feature clusters in the current frame of image data using identifiers associated with the feature cluster in the previous frame of image data that have been correlated with the feature clusters in the current frame of image data.

Abstract: A computer-automated method for detecting a vessel in water based on an image of a portion of Earth includes generating a thermal anomaly mask. The thermal anomaly mask flags each pixel of the image initially deemed to be a wake pixel based on a comparison of a thermal value of each pixel against other thermal values of other pixels localized about each pixel. Contiguous pixels flagged by the thermal anomaly mask are grouped into pixel clusters. A shape of each of the pixel clusters is analyzed to determine whether each of the pixel clusters represents a possible vessel detection event. The possible vessel detection events are represented visually within the image.

Abstract: A method of automatically identifying bone components in a medical image data set of voxels, the method comprising: a) applying a first set of one or more tests to accept voxels as belonging to seeds, wherein none of the tests examine an extent to which the image density has a local maximum at or near a voxel and falls steeply going away from the local maximum in both directions along an axis; b) applying a second set of one or more tests to accept seeds as bone seeds, at least one of the tests requiring at least one voxel belonging to the seed to have a local maximum in image density at or near said voxel, with the image density falling sufficiently steeply in both directions along at least one axis; and c) expanding the bone seeds into bone components by progressively identifying candidate bone voxels, adjacent to the bone seeds or to other previously identified bone voxels, as bone voxels, responsive to predetermined criteria which distinguish bone voxels from voxels of other body tissue.

Abstract: A method for detecting, scaling, and positioning handwritten information from hardcopy media, the method includes the steps of digitizing the media and detecting handwritten information; analyzing the media to determine an available location in the digitized image for placement; determining the orientation of the media and handwritten information; and orienting, scaling, and positioning the detected handwritten information. The method further includes storing or using the digitized image side with the handwritten information.

Abstract: A system for generating virtual interest points in an image, according to detected interest points in that image is detected by an interest point detector. The system includes: an interest point density map producer and an interest point generator, the interest point density map producer receiving the image including the detected interest points, the interest point density map producer extracts a density map of the detected interest points within the image. The interest point generator is coupled with the interest point density map producer, the interest point generator receiving the image including the detected interest points and the interest point density map. The interest point generator determines image areas for generating virtual interest points, the interest point generator generating at least one virtual interest point at the determined areas according to at least the position of one of the detected interest points, the virtual interest point not being detected by the interest point detector.

Abstract: Provided is a microscope apparatus including a macro-image, acquisition section that acquires a macro image of a glass slide having a specimen mounted thereon; an extraction section that extracts the specimen in the acquired macro image; a block setting section that sets, when the extracted specimen is scattered to form a plurality of lumps, a plurality of blocks that include the lumps of the specimen; an area dividing section that divides an area that includes the specimen in each of the blocks into a plurality of small regions; and a micro-image acquisition section that acquires, for each of the small regions, a micro image while performing an automatic focusing operation for the specimen in the small region, in which the micro-image acquisition section searches for an autofocusing in-focus position in each of the blocks set by the block setting section.

Abstract: A text-like data representation technique and a text-like data representation apparatus are disclosed that may: acquire image data from a scanned image; segment text regions from the image data; further extract each connected component in the text regions; form clusters based on the connected components; group each connected component in the text regions into one of the clusters with similar or identical characters; generate a high-resolution representative for each cluster; generate a vector representation for each high-resolution representative; and code the text as text data by associating each connected component with its vectorized high-resolution representative, and location in the document.

Abstract: A protection method for preventing a hard copy of a document from being released or reproduced is provided. The protection method comprises: selecting a pattern with a closed contour line and utilizing a level set method to process the pattern to generate at least a watermark; and embedding the watermark into the document. In addition, the protection method can further comprise: detecting the hard copy of the document to determine whether at least a watermark exists in the hard copy of the document; and forbidding releasing or reproducing the document when the watermark is detected in the hard copy of the document.

Abstract: A method and system generates fine-grained fingerprints for identifying content in a rendered document. It includes applying image-based techniques to identify patterns in a document rendered by an electronic document rendering system, irrespective of a file format in which the rendered document was electronically created. The applying of the image-based technique includes identifying candidate keypoints at locations in a local image neighborhood of the document, and combining the locations of the candidate keypoints to form a fine-grained fingerprint identifying patterns representing content in the document.

Abstract: A method of image processing includes: receiving at least one video frame of a video sequence, the at least one video frame including at least one foreground subject and a background; and processing the at least one video frame so as to separate the at least one foreground subject from the background. The processing includes: generating a pixel mask containing information indicating, for each pixel of the at least one video frame, whether the pixel belongs to the foreground subject or to the background, and determining contours of the at least one foreground subject on the pixel mask. The determining of the contours includes for each pixel in the at least one video frame; based on the information included in the pixel mask, determining whether at least one pixel border belongs to a contour of the at least one foreground subject, the at least one pixel border separating the pixel from a respective at least one adjacent pixel.

Abstract: Methods and systems detect objects of interest in a spatio-temporal signal. According to one embodiment, a system processes a digital spatio-temporal input signal containing zero or more foreground objects of interest superimposed on a background. The system comprises a foreground/background separation module, a foreground object grouping module, an object classification module, and a feedback connection. The foreground/background separation module receives the spatio-temporal input signal and, according to one or more adaptable parameters, produces foreground/background labels designating elements of the spatio-temporal input signal as either foreground or background. The foreground object grouping module is connected to the foreground/background separation module and identifies groups of selected foreground-labeled elements as foreground objects. The object classification module is connected to the foreground object grouping module and generates object-level information related to the foreground object.

Abstract: A portion of an image that has been compressed is displayed promptly. To achieve this, an image is divided into a number of units and an RST marker is inserted into a unit every ten units. The image is compressed and recorded unit by unit. In a case where an area which is a portion of the image is displayed, the image data of all units of the original data is not expanded. Rather, units in which the image data is to be expanded are decided utilizing the RST markers. Only the image data of the units decided is expanded.

Abstract: A method of skin segmentation of a digital image is operable in an acquisition device. An image is acquired. A value indicative of a redness of a pixel of said image is compared with a face skin pixel redness criterion. The pixel is identified as a face skin pixel if said criterion is satisfied.

Abstract: Disclosed is a print controlling device including a control unit to control printing according to an instruction which relates to the printing, wherein the control unit analyzes a page description language which relates to the printing, determines whether the printing is practically monochrome printing or color printing, and changes a first command of the page description language to a second command for monochrome printing when it is determined that the printing is monochrome printing.

Abstract: Computer graphics may be detected in digital images by extracting a first set of features from an input digital image, extracting a second set of features from a prediction-error image derived from the input digital image, and applying a classification algorithm to the first set of features and the second set of features to determine if the combined sets of features indicate that the input digital image corresponds to computer graphics.

Abstract: A method, apparatus, and system, for scanning a first portion of a data to generate a second portion of data is provided. A control parameter relating to a level of detail associated with filtering a first portion of data is received. The filtering of the first portion of data is performed based upon the control parameter. The filtering of the first portion of data includes a rule-based filtering, a context-based filtering, a statistical-based filtering, or a semantic-based filtering. Performing the filtering provides for a reduction of a portion of the first portion of data. A second portion of data that is smaller than the first portion of data is provided based upon the filtering of the first portion of data.

Abstract: In a method and apparatus for determining a classification boundary between an object, such as a vehicle, and a background, using an object classifier, a trained classifier is configured to classify and recognize each of a plurality of object images and a plurality of background images. Next, a confidence probability density distribution function is calculated for the vehicle and the background using the determined confidence values for the vehicle images and background images. Once the probability density distribution functions for the vehicle and the background are calculated, the classification boundary between the vehicle and the background is determined using the probability density distribution functions for the vehicle or the background, or both, and a predefined model that is appropriate.

Abstract: A system and method of labeling orthogonal or otherwise spatially related image views and related images is provided. The present invention provides automated progression for the labeling of vertebral and inter-vertebral regions, propagation of labels between views and images within a series, centering of label regions relative to the spine, circular lists of predefined labels, and label displays for individual slices of an orthogonal or axial view as a user scrolls through the plurality of image slices of the given view. In a further aspect, the present invention provides automated labeling of vertebral and inter-vertebral regions when a user provides labels for the adjacent two inter-vertebral or vertebral regions.

Abstract: A method of generating a depth map for an image comprises the following steps: displaying the image, displaying an indication on the image in a first region of the image, displaying a plurality of labels, receiving a user input selecting one of the plurality of labels, setting a label for the first region to the selected label, repeating steps two to five for a different region, until a predefined event has occurred, accessing a predefined scheme for converting a label map to a depth map, and generating a depth map according to the predefined scheme and labels set following the user inputs.

Abstract: In a method of displaying an image from an image store to a display device, each pixel has a color represented by pixel values which define a position within color space. A plurality of regions are defined within the color space, and each pixel value used within the image is allocated to one of these regions within the color space. Then, one or more areas of the image are selected, and for each of the selected areas, one of the defined regions within the color space is selected. For each selected area of the image, pixels which have been allocated to the selected region within the color space are identified and added to the image store. Different areas of the image and regions of the color space are selected in succession until all pixels of the image have been added to the image store for display.

Abstract: A method for segmentation of locally-defined regions is provided. The input data obtained for the image region under investigation is separated into a foreground mask and a background mask, two different filter operations are applied to the input data to determine segmentation starting points, the resulting data sets are subtracted and the image regions viewed as segmentation starting points, a filter operation is applied to generate a less restrictive mask for the input data, the resulting data subjected to a threshold value reduction, the less restrictive mask subjected to a labeling with regard to determination of the local values for generation of a restrictive mask, the mean values determined, segmentation starting points not obtained in the less restrictive mask are removed, a growth into all regions of the restrictive mask is carried out and the regions obtained in the restrictive mask are viewed as the segmented brightness regions.

Abstract: An image processing method and image processing apparatus which can execute vector conversion processing by appropriately dividing a clipart image including gradation into regions are provided. To this end, a color document image including a clipart image having an area of gradation is input, the clipart image is selected, and outline information of the clipart image is extracted. The color document image is separated into an achromatic color area and a chromatic color area, which are respectively divided into a plurality of regions. Of the plurality of regions of each of the achromatic color area and the chromatic color area divided in the dividing of region step, regions which meet a set condition are integrated. After that, the clipart image is converted into vector data using a region group after integrating of region and the outline information.

Abstract: An information storage medium including graphic data and presentation information, and an apparatus and method of processing the graphic data are provided. The information storage medium includes the graphic data, page composition information which defines page composition of the graphic data, and the presentation information indicating when graphic screen data, which is composed with reference to the page composition information of the graphic data, is output to a display screen. Therefore, a graphic object is reusable in graphic data processing, and accordingly, a time taken to process the graphic data is reducible and memory area may be saved.

Abstract: An Internet/web page printer printing other web pages linked to a printed web page and a method thereof. A URL hyperlinked to the printed web page is extracted from web page information, which is obtained through a network via a web browser and converted into image information to be printed. From the converted image information, URL region information is obtained indicating a print region in which a URL anchor corresponding to the extracted hyperlinked URL is located on the printed web page. Location information input through a location input unit corresponding to the printed image information is compared with the URL region information to select an extracted hyperlinked URL corresponding to the input location information. The selected extracted hyperlinked URL is provided to the web browser so that corresponding web page information is obtained, thereby obtaining other web pages hyperlinked to the printed web page.

Abstract: A method and system analyzes data acquired by image systems to more rapidly identify objects of interest in the data. In one embodiment, z-depth data are segmented such that neighboring image pixels having similar z-depths are given a common label. Blobs, or groups of pixels with a same label, may be defined to correspond to different objects. Blobs preferably are modeled as primitives to more rapidly identify objects in the acquired image. In some embodiments, a modified connected component analysis is carried out where image pixels are pre-grouped into regions of different depth values preferably using a depth value histogram. The histogram is divided into regions and image cluster centers are determined. A depth group value image containing blobs is obtained, with each pixel being assigned to one of the depth groups.

Abstract: In an image-processing apparatus having a capability of performing region distinction processing and an image region discrimination processing method, a first region distinction unit uses a previously set threshold value for an image region distinction to perform a region distinction processing of a character and a non-character on image data read from an original document, an edge feature amount image and a character determination signal are obtained, a second region distinction unit makes a region distinction on the edge feature amount image based on the threshold value and generates and displays sub-region images obtained by dividing the edge feature amount image into plural parts, a character discrimination strength adjustment is performed on a display screen while each of the sub-region images is visually identified, the correction parameter is reflected in the edge feature amount image, and the region distinction processing is performed again.

Abstract: A method exploits user labels in image segmentation. The user labels are propagated with respect to image intensity information. Propagated user labels are included in a cost function of level set evolution. The level set represents a probability of the object segment.

Abstract: A document authentication method compares a target document image (scanned image) with an original document image at multiple levels, such as block (e.g. paragraph, graphics, image), line, word and character levels. The paragraph level comparison determines whether the target and original images have the same number of paragraphs and whether the paragraphs have the same sizes and locations; the line level comparison determines if the target and original images have the same number of lines and whether the lines have the same sizes and locations; etc. Document segmentation is performed on the target and original images to segment them into paragraph units, line units, etc. for purposes of the comparisons. The original document may be segmented beforehand and the segmentation information stored for later use. The authentication process may be designed to stop when alterations are detected at a higher level, so lower level comparisons are not carried out.

Abstract: One of the significant problems encountered in criminology studies is the successful automatic matching of cartridge cases after a cartridge is fired from a firearm, on the basis of the marks left on the cartridge cases fired. One of the probable steps in the solution of this problem is the segmentation of certain regions defined on the cartridge case. This invention relates to the method for segmentation of cartridge case base by using surface height matrix information.

Abstract: The class of a sheet is efficiently estimated and a pattern identification process which is robust to a variation in the medium can be performed by dividing an image pattern of the sheet into a plurality of areas (pixels or sets of pixels), weighting and selecting the areas, attaining the identification results for the respective areas and determining the identification result of the whole portion based on a logical combination of the identification results. Particularly, since the area weighting and selecting process is performed based on a difference between the classes and a variation in the class, the calculation amount can be reduced and the identification performance which is higher than that of a method which uniformly processes the whole portion of the pattern can be attained.

Abstract: Real-time segmentation of foreground from background layers in binocular video sequences may be provided by a segmentation process which may be based on one or more factors including likelihoods for stereo-matching, color, and optionally contrast, which may be fused to infer foreground and/or background layers accurately and efficiently. In one example, the stereo image may be segmented into foreground, background, and/or occluded regions using stereo disparities. The stereo-match likelihood may be fused with a contrast sensitive color model that is initialized or learned from training data. Segmentation may then be solved by an optimization algorithm such as dynamic programming or graph cut. In a second example, the stereo-match likelihood may be marginalized over foreground and background hypotheses, and fused with a contrast-sensitive color model that is initialized or learned from training data. Segmentation may then be solved by an optimization algorithm such as a binary graph cut.

Abstract: An image processing apparatus for extracting pictures-in-picture information contained in an image, includes an image segmentation unit segmenting an input image into regions, and a region extraction unit extracting a pictures-in-picture region containing the pictures-in-picture information on the basis of a contour of a segmented region.

Abstract: A gamma imaging device including a gamma camera capturing a gamma radiation image, referred to as a gamma image, of an observed scene, and including a front face and having an axis of sight, and an auxiliary camera capturing a visible light image of the observed scene. The auxiliary camera is situated upstream from the front face of the gamma camera and has an optical axis substantially merged with the axis of sight of the gamma camera such that the visible light image and the gamma image are captured substantially simultaneously with the same direction of sight.

Abstract: In an intelligent video surveillance system, video processing software performs a number of operations on video data received from a camera, including foreground extraction, shadow removal and object tracking. The foreground extraction stage classifies each pixel of a received frame as representing either foreground or background. Since shadow regions can be wrongly classified as foreground, a two-branch shadow removal operation is employed, comprising weak shadow removal (32) and strong shadow removal (33). The purpose of weak shadow removal (32) is to remove only the most severe shadows in each frame. Conversely, the purpose of strong shadow removal (33) is to remove substantially every shadow present in each frame. By comparing the overlap of foreground regions in the two resulting images, it is possible to identify foreground regions that have fragmented due to strong shadow removal, and thereafter classify them as representing a common foreground object.

Abstract: To identify pixels constituting image elements for generating a labeled image in which pixels are labeled with identification information, a pixel block including four pixels adjacent to one another in two dimensions is inputted as a unit from data including pixels that form an image. All of the on-pixels that are subject for grouping, included in the pixel block are labeled with the same identifier. Since the on-pixels that are included in a pixel block are consecutively connected, such pixels can be labeled with the same identifier without having to calculate whether such pixels are connected.

Abstract: A method of labeling pixels in an image is described where the pixel label is selected from a set of three or more labels. The pixel labeling problem is reduced to a sequence of binary optimizations by representing the label value for each pixel as a binary word and then optimizing the value of each bit within the word, starting with the most significant bit. Data which has been learned from one or more training images is used in the optimization to provide information about the less significant bits within the word.

Abstract: A sensing device for enabling a user to interact with a computer system. The sensing device is operable in a cursor control mode and a substrate-interaction mode. The sensing device includes: a nib; a nib switch configured to effect a mode change between the cursor control mode and the substrate-interaction mode; an optical sensor for sensing coded data printed on a substrate; a processor for generating cursor control data or interaction data; and a transceiver for communicating the cursor control data or the interaction data to a computer system. The cursor control data initiates a cursor control response in the computer system, while the interaction data initiates a substrate interaction response in the computer system.

Abstract: To identify pixels constituting image elements for generating a labeled image in which pixels are labeled with identification information, a pixel block including four pixels adjacent to one another in two dimensions is inputted as a unit from data including pixels that form an image. All of the on-pixels that are subject for grouping, included in the pixel block are labeled with the same identifier. Since the on-pixels that are included in a pixel block are consecutively connected, such pixels can be labeled with the same identifier without having to calculate whether such pixels are connected.

Abstract: A method for labeling connected components and a computer system using the method are provided. With the method, during a process of scanning an image for the first time, each object pixel in the image is assigned a temporary label, and a relationship between each temporary label and a representative label is established after the completion of the process of scanning the image for the first time. Thereafter, the temporary label of each object pixel is replaced by the corresponding representative label during a process of scanning the image for the second time. As a result, the labeling of the connected components can be accomplished by only scanning the image twice such that the efficiency of labeling the connected components can be significantly improved.

Abstract: Image data which includes a feature is labeled. The image data is comprised of pixel data in an N×M array. At least one pixel is designated and flagged. Asynchronous processes are executed, where each process corresponds to exactly one pixel. Each process sets a flag only for the owner pixel and only if a connected neighbor pixel in the pre-defined path is flagged. Each process inspects pixel data for all neighbor pixels of the owner pixel to determine if a neighbor pixel is a connected pixel and if the neighbor pixel is flagged, and continues until it sets the flag for the owner pixel or until all connected horizontal and vertical neighbor pixels have been inspected without encountering a connected neighbor pixel whose flag is set. The asynchronous processes are repeated until a pre-defined condition has been satisfied. All flagged pixels are labeled.

Abstract: Methods for dimensionality reduction of large data volumes, in particular hyper-spectral data cubes, include providing a dataset ? of data points given as vectors, building a weighted graph G on ? with a weight function w?, wherein w? corresponds to a local coordinate-wise similarity between the coordinates in ?; obtaining eigenvectors of a matrix derived from graph G and weight function w?, and projecting the data points in ? onto the eigenvectors to obtain a set of projection values ?B for each data point, whereby ?B represents coordinates in a reduced space. In one embodiment, the matrix is constructed through the dividing each element of w? by a square sum of its row multiplied by a square sum of its column. In another embodiment the matrix is constructed through a random walk on graph G via a Markov transition matrix P, which is derived from w?. The reduced space coordinates are advantageously used to rapidly and efficiently perform segmentation and clustering.

Abstract: Systems and methods are disclosed to perform image parsing on one or more images by identifying a set of similar regions from each image; assigning one or more region labels to each region and generating multiple hypotheses for region label assignment; and detecting class, location and boundary of each object in the image, wherein object classification, detection and segmentation are performed jointly during image parsing.

Abstract: Methods and systems for cropping images of book pages are disclosed according to one embodiment of the invention. A method may include identifying reference images and receiving cropping rectangles for the reference images. These cropping rectangles associated with reference images may then be used to generate cropping rectangles for images of book pages between the reference images. The cropping rectangles may be generated based on a linear interpolation of the cropping rectangles associated with the reference images and the number of pages between images. The method may also display one or more images of book pages with the associated one or more cropping rectangles superimposed thereon. A user may then have the opportunity to make adjustments to the position and/or size of the cropping rectangles.

Abstract: The present invention discloses a multi-point image labeling method, which labels an object pixel matrix containing image data and makes adjacent array elements with image data have an identical image label value. A multi-point label window is used to designate a non-zero temporary labeled value to store in the register according to the temporary labeled value of the adjacent array elements. Next, a label-equivalence window generates label-equivalence information according to the adjacent temporary labeled values. Next, an equivalent-substitution processing process is performed on the temporary labeled values according to the label-equivalence information to generate label-equivalence substitution information. Then, the temporary labeled values are replaced according to the label-equivalence substitution information to obtain the resultant image labeled values and complete the image labeling of the object pixel matrix.

Abstract: A method and system are described for re-arranging a display in a viewable area. The method comprises dividing the viewable area into units; determining the contents of units of the viewable area; prioritizing the availability of the units by content; and allocating one or more display items in the prioritized available units. A plurality of display items is provided and the display items are allocated according to a determined order which may depend on one or more of a user preference, the contents of the viewable area, the activity of the user, and environmental parameters.

Abstract: Methods for detecting a salient object in an input image are described. For this, the salient object in an image may be defined using a set of local, regional, and global features including multi-scale contrast, center-surround histogram, and color spatial distribution. These features are optimally combined through conditional random field learning. The learned conditional random field is then used to locate the salient object in the image. The methods can also use image segmentation, where the salient object is separated from the image background.