Abstract: Methods and apparatus to generate templates from web images for searching an image database are described. In one embodiment, one or more retrieved images (e.g., from the Web) may be used to generate one or more templates. The templates may be used to search an image database based on features commonly shared between sub-images of the retrieved images. Other embodiments are also described.

Abstract: A method of multi-frame image noise reduction suitable for an image-capturing device includes following steps: obtaining a current frame and multiple reference frames; defining a mask and a target point in the mask; judging whether the target point pixel of the current frame is on an edge according to an edge map of the frame; when the pixel is on the edge, using the pixels in the reference frames on the edge to calculate a replacement result; when the target point pixel is not on the edge, using the pixels in the reference frames surrounding the target point to calculate a replacement result; after that, generating a pixel corresponding to the position of the target point in an output image according to the replacement result; further, moving the mask and going back to judge whether the pixel of the target point of the current frame is on the edge.

Abstract: An image processing apparatus for searching for corresponding points between a first image and a second image is provided. The image processing apparatus includes a window setting portion, a phase information extraction portion that extracts phase information by subjecting a representation included in each set window to frequency decomposition, a corresponding points determination portion that determines the corresponding points based on comparison of the phase information extracted from each window, and a window size control portion that causes the window setting portion to set a window different in size in accordance with a frequency used for calculation of the frequency decomposition. The window size control portion causes the window setting portion to set a window having a first size in response to a first frequency and to set a window having a second size smaller than the first size in response to a second frequency higher than the first frequency.

Abstract: One or more aspects of the subject disclosure are directed towards identifying objects within an image via image searching/matching. In one aspect, an image is processed into bounding boxes, with the bounding boxes further processed to each surround a possible object. A sub-image of pixels corresponding to the bounding box is featurized for matching with tagged database images. The information (tags) associated with any matched images is processed to identify/categorize the sub-image and thus the object corresponding thereto.

Abstract: A method for identifying the connectivity of texture types represented in a digital image comprising pixels, each pixel having a texture value which is representative of texture at a respective position, the method comprising: partitioning the texture values into local neighborhoods of texture values; determining a directionality for each neighborhood; using the directionality of the neighborhoods to determine their nearest neighborhood or neighborhoods; connecting the neighborhoods with their nearest neighborhood or neighborhoods; and determining the connectivity of the texture types of the digital image based on the connections formed between neighborhoods.

Abstract: Systems and methods for object detection are presented herein. Embodiments of the present invention utilizing a cascade feature, one or more features at different scales, one or more multi-scale features in combination with a perspective feature, or combinations thereof to detect an object of interest in an input image. In embodiments, the various features are used to train classifiers. In embodiments, the trained classifiers are used in detecting an object of interest in one or more input images.

Abstract: An image processing apparatus for extracting a contour of an object which extracts boundary candidate points from the image and sets reference areas each having one of boundary candidate points as a center and includes small areas. A first reference area is set from the reference areas and a first small area is set from small areas included in the first reference area. From reference areas, a second reference area is set, whose feature is similar to that of the first reference area and as a second small area, a small area is set which has a feature amount similar to that of the first small area among small areas included in the second reference area. The boundary candidate points is tracked to extract the contour of the object based on the first reference area, the second reference area, the first small area, and the second small area.

Abstract: A method for detecting a text region in an image is disclosed. The method includes detecting a candidate text region from an input image. A set of oriented gradient images is generated from the candidate text region, and one or more detection window images of the candidate text region are captured. A sum of oriented gradients is then calculated for a region in one of the oriented gradient images. It is classified whether each detection window image contains text by comparing the associated sum of oriented gradients and a threshold. Based on the classifications of the detection window images, it is determined whether the candidate text region is a true text region.

Abstract: An analysis of a digitized image is provided. The digitized image is repeatedly convolved to form first convolved images, which first convolved images are convolved a second time to form second convolved images. Each first convolved image and the respective second convolved image representing a stage, and each stage represents a different scale or size of anomaly. As an example, the first convolution may utilize a Gaussian convolver, and the second convolution may utilize a Laplacian convolver, but other convolvers may be used. The second convolved image from a current stage and the first convolved image from a previous stage are used with a neighborhood median determined from the second convolved image from the current stage by a peak detector to detect peaks, or possible anomalies for that particular scale.

Abstract: Systems, methods and computer-readable storage media are disclosed for accelerating bitmap remoting by extracting non-grid tiles from source bitmaps. A server takes a source image, identifies possibly repetitive features, and tiles the image. For each tile that contains part of a possibly repetitive feature, the server replaces that part with the dominant color of the tile. The system then sends to a client a combination of new tiles and features, and indications to tiles and features that the client has previously received and stored, along with an indication of how to recreate the image based on the tiles and features.

Abstract: It is determined whether or not an input image is an image converted from an image with a relatively low resolution based on one frame of an image. A resolution determination device includes: an edge strength calculator configured to obtain an edge strength of a pixel included in an input image based on luminance of the pixel and luminance of a pixel adjacent to the pixel, for each of a plurality of pixels included in the input image; and a resolution determiner configured to determine whether or not the input image is an image upconverted from an image with a predetermined resolution or less, based on distribution of the edge strengths.

Abstract: Method, system and computer program for efficiently extracting and describing scale-space interest points. It is designed towards low overall computational complexity. On one hand, the data acquired during extraction in the description phase is intensively re-used. On the other hand, an algorithmic optimization of the description that dramatically speeds up the process, is proposed. First, the image is filtered with triangle kernel at different scales. The triangle filtered images are reused for extraction of the keypoints dominant orientation and the computation of the DAISY-like descriptor.

Abstract: A method for processing an image is described. Mask bits are determined for a current pixel. The mask bits indicate intensity comparisons between the current pixel and multiple neighboring pixels. The mask bits also indicate whether each of the current pixel's neighboring pixels have been processed. A next pixel is selected for processing based on the mask bits.

Abstract: According to one embodiment, an image processing apparatus includes following units. The selection unit selects a pixel from a target image. The first extraction unit extracts a first pixel region including the selected pixel. The second extraction unit extracts a second pixel region from a reference image. The determination unit determines a filter coefficient based on a similarity degree between the first and second pixel regions. The generation unit generates a display image by performing a weighted sum of the target image and a display image generated immediately before the target image, in accordance with the filter coefficient determined for each of the plurality of pixels of the target image.

Abstract: Systems, methods and computer-readable storage media are disclosed for accelerating bitmap remoting by extracting non-grid tiles from source bitmaps. A server takes a source image, identifies possibly repetitive features, and tiles the image. For each tile that contains part of a possibly repetitive feature, the server replaces that part with the dominant color of the tile. The system then sends to a client a combination of new tiles and features, and indications to tiles and features that the client has previously received and stored, along with an indication of how to recreate the image based on the tiles and features.

Abstract: Applications, systems and methods for identifying and monitoring critical colors in a print job and using an embedded color sensing device to measure critical color printed actually printed. The application permits the user to interact with a print job and to select of one or more critical colors from within the print job. In one embodiment, an application is configured to provide a graphical user interface (GUI) to facilitate the user in selecting critical colors. A printing system and method are disclosed for automatically monitoring critical colors in a print job and using an embedded color sensing device to measure critical color printed.

Abstract: An approach to computation of kernel descriptors is accelerated using precomputed tables. In one aspect, a fast algorithm for kernel descriptor computation that takes O(1) operations per pixel in each patch, based on pre-computed kernel values. This speeds up the kernel descriptor features under consideration, to levels that are comparable with D-SIFT and color SIFT, and two orders of magnitude faster than STIP and HoG3D. In some examples, kernel descriptors are applied to extract gradient, flow and texture based features for video analysis. In tests of the approach on a large database of internet videos used in the TRECVID MED 2011 evaluations, the flow based kernel descriptors are up to two orders of magnitude faster than STIP and HoG3D, and also produce significant performance improvements. Further, using features from multiple color planes produces small but consistent gains.

Abstract: In an embodiment, a device comprises a plurality of elements configured to apply a filter to multiple groups of pixels in a neighborhood of pixels surrounding a particular pixel to generate a matrix of filtered values; compute, from the matrix of filtered values, a first set of gradients along a first direction and a second set of gradients along a second and different direction; determine how many directional changes are experienced by the gradients in the first set of gradients and the gradients in the second set of gradients; compute a first weighted value for a first direction and a second weighted value for a second direction; and based, at least in part, upon the first and second weighted values, compute an overall texture characterization value for the particular pixel, wherein the overall texture characterization value indicates a type of image environment in which the particular pixel is located.

Abstract: An automated printout inspection system identifies glyphs in an image by calculating a connectedness score for each foreground pixel, and comparing this score with a specified threshold. The system further generates training images by simulating printouts from an impact printer, including the specifying of specific error types and their magnitudes. The simulated printouts are combined with scan images of real-world printout to train an automated printout inspection system. The inspection results of the automated system are compared with inspection results from human inspectors, and test parameters of the automated system are adjusted so that it renders inspection results within a specified range of the average human inspector.

Abstract: A method for extracting an epitome from an image Y is described. The method comprises: searching for each block, a list of patches in the image Y with similar content; determining, for each patch of said lists, the set of blocks able to be represented by the patch; constructing, from the sets of blocks, epitome charts comprising, for each epitome chart, initializing said epitome chart with one patch and extending said epitome chart with at least one candidate region contained in a patch overlapping said epitome chart; said method being characterized in that said candidate region is selected by taking into account the blocks represented by said patch overlapping said epitome chart and by any other patch contained in said epitome chart enlarged by said selected candidate region.

Abstract: A pedestrian detector includes an image-taking mechanism and an image analyzer. The image-taking mechanism is provided on a vehicle and used to take an image in front of the vehicle. The image analyzer is connected to the image-taking mechanism and used to receive the image from the image-taking mechanism and analyze the image to determine whether the image conforms to a pedestrian to determine whether there is a pedestrian in front of the vehicle.

Abstract: An image processing apparatus identifies tissues in respective parts of a tissue image. A tissue image subdivider subdivides a tissue image for identification into local regions. A detector detects texture feature values of the local regions. A determining unit compares the detected texture feature value of a local region to a learned feature value for identification associated with a predetermined tissue, and on the basis of the comparison result, determines whether or not the local region belongs to the predetermined tissue.

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: The invention relates to an improved robustness in upscaling the resolution of regularly sampled multi-dimensional signals, where a single low-resolution signal is available. These methods are referred to as example-based super-resolution or single-image super-resolution. A method for super-resolving a single image comprises three stages. First, an interpolation-based up-scaling of the input image is performed, followed by a Local De-noising step in contour regions. The second stage comprises extrapolation through cross-scale block matching, wherein an extrapolated high-frequency band is obtained that is de-noised through regularization in the same contour regions. The third stage comprises adding the contributions of the low-frequency band of the high-resolution image and the extrapolated high-frequency band.

Abstract: In an embodiment, a device comprises a plurality of elements, including logical elements, wherein the elements are configured to perform the operations of: in a neighborhood of pixels surrounding and including a particular pixel, applying a filter to multiple groups of pixels in the neighborhood to generate a set of filtered values; generating, based at least in part upon the set of filtered values, one or more sets of gradient values; based at least in part upon the one or more sets of gradient values, computing a first metric for an image environment in which the particular pixel is situated; determining a second metric for the image environment in which the particular pixel is situated, wherein the second metric distinguishes between a detail environment; and based at least in part upon the first metric and the second metric, computing a gradient improvement (GI) metric for the particular pixel.

Abstract: A spot quantification apparatus includes an image input unit, a filtering unit, and a spot quantification unit. The image input unit receives an image including a spot. The filtering unit filters the image input through the image input unit a plurality of times while changing a parameter of a kernel within a predetermined range. The spot quantification unit quantifies a feature of the spot based on the luminance distribution in each image which has been filtered a plurality of times by the filtering unit.

Abstract: Some of the embodiments of the present disclosure provide a method comprising selecting a pixel window of image data, the pixel window including a target pixel, determining a stability of the pixel window, formulating a look up table address based at least in part on the determined stability, obtaining one or more image enhancement values from a look up table based at least in part on the formulated look up table address, and processing the target pixel based at least in part on the obtained one or more image enhancement values. Other embodiments are also described and claimed.

Abstract: An image processing device comprises a store (110) for receiving image data (8) with a pixel identifier and at least one piece of associated intensity data and a pilot (100) which is provided to call a reduction unit (120) with pixels to be weighted and, for each pixel, in order to define a first neighborhood and a plurality of second neighborhoods with a neighborhood selection unit (130), in order to obtain a weighting value for the second neighborhood with a weighting unit (140) with the first neighborhood and a neighborhood among the plurality of second neighborhoods, in order to define each time a piece of weighted intensity data of a working pixel as the weighting value of the second neighborhood multiplied by the intensity data of the associated working pixel, and in order to define a piece of weighted intensity data of the pixel to be weighted by adding each of the pieces of weighted intensity data of a working pixel.

Abstract: A method and system for analyzing an image made up of a plurality of pixels includes identifying a subset of the pixels, and for each identified pixel a local patch of pixels surrounding the identified pixel is defined and the pixels in the local patch are grouped into bins. The local patch is divided into a plurality of sub-patches, and for each sub-patch the method includes determining how many pixels in each sub-patch fall into each of the bins to create an intermediate vector, and the intermediate vectors are concatenated to form a feature vector and describe the local patch.

Abstract: A method for identifying singleton outlier pixels in a selected color space in a digital image including a plurality of pixels, includes for each 3×3 patch of pixels in the image, calculating the diameter of the 3×3 patch of pixels. For each pixel in the patch, the distance to its nearest neighbor pixel within the patch is computed, as measured in the selected color space. The computed distance from each pixel in the patch is compared to its nearest neighbor with a threshold that is a preselected fraction of the diameter. A center pixel in the patch is identified as an outlier pixel if its calculated distance to its nearest neighbor is the largest distance to a nearest neighbor and exceeds the threshold.

Abstract: A method and system of digital image enhancement provide controllable emboldening and thinning of image features. The method includes providing an input image, creating an edge map image from the input image, and combining the edge map image with the input image using local clamping to produce an output image. Features of the output image are one of emboldened and thinned.

Abstract: The present invention relates to an image processing apparatus for compressing image data used in an image generating apparatus for generating a free-viewpoint image. According to the invention, the apparatus has a selecting unit that selects one image as a first image, and defines other images as second images, a projective transformation unit that generates a projected depth map of a second image from a depth map of the first image, a subtracting unit that creates a difference map of the second image, and a storage unit that stores the depth map of the first image and the difference map of the second image. Here, the difference map is a difference between a depth map of the second image and the projected depth map of the second image, and the depth map indicates a depth value of each pixel of a corresponding image.

Abstract: An image retrieval device that retrieves an image in response to an input image by comparing feature amounts based on an attribute of search target images and a feature amount based on an attribute of an input image. Each of the feature amounts comprises a first feature amount expressing a global relative positional relation of each pixel constituting an image with respect to pixels of the entire image, and a second feature amount expressing a local relative positional relation of each pixel constituting each of the images with respect to its adjacent pixels. Component images and at least one combination image formed therefrom are linked to a search target image and feature amounts about these respective images are compared with a feature amount about an input image and a search target image linked to a component/combination image that is similar to the input image is displayed.

Abstract: An image processing apparatus for correcting a positional deviation between a plurality of images obtained in time series and performing synthesis processing on the plurality of corrected images determines a motion vector between the plurality of images, corrects the positional deviation between images on the basis of the determined motion vector, and generates a synthesized image by synthesizing the plurality of images subjected to the positional deviation correction. The image processing apparatus extracts a filter area in which filter processing is to be performed on the basis of a degree of inconsistency occurring when the positional deviation between the plurality of images is corrected, determines a filter kernel on the basis of an imaging condition, and performs filter processing based on the filter kernel on the filter area of the synthesized image.

Abstract: An image processing apparatus including: a determining processing section which binarizes a pixel value of a target pixel using a threshold; a calculation section which calculates feedback values by multiplying binarized pixel values of adjacent processed pixels by a feedback coefficient; a addition section which adds the feedback values to the pixel value of the target pixel before binarization; a diffusion section which diffuses a difference between the binarized pixel value and the unadded pixel value as an error to adjacent unprocessed pixels; and a subtraction section which subtracts the error from the unadded pixel value of the target pixel, wherein when all pixel values of processed pixels are 0 and an output value of the target pixel is a maximum value, it is determined that the target pixel is an isolated point and an output value of an adjacent pixel is set to a forced output value excluding 0.

Abstract: We describe a method and apparatus for monitoring the dripping speed in IV process, as well as detecting the its end. The apparatus includes a camera and a processing unit to analyze the information from the acquired video. Features from the image sequence are extracted and the dripping speed is computed accordingly by discrete Fourier transform. The apparatus is also capable of detecting the end of the dripping process by finding the location of liquid surface in the drip chamber. We also describe the use of barcode to provide information to the monitoring device and program.

Abstract: A method of creating a container file for large format imagery and organizing data within the container file are described. In one embodiment, the method of creating the container file includes validating user input parameters for the file container, and determining whether the container file already exists, the container file having file container metadata. If the file container does not exist, creating the container by creating one or more empty records in a storage device, the one or more empty records having an image file section reserved for storing an image, an image metadata section reserved for storing data about the image and a record metadata section having at least a mark indicating that the image file section is empty. A size of the image file section, a size of the image metadata section and a size of the record metadata section are determined using the user input parameters.

Abstract: The present invention enables mixture of a core line vectorization process and a outline vectorization process, and comprises: inputting an object image of a vectorization process; extracting a core line of the object image; computing an original line width for every pixel of the core line acquired by the extracting; judging whether every section is indicative of an equal-width line or indicative of a surface figure by using the line width value for every pixel of the core line acquired by the estimating, the every unit delimiting the core line acquired by the extracting between two points of intersections and endpoints; separating a surface figure from the object image on the basis of a judging result of the judging; and approximating functionally the core line judged to be the equal-width line at the judging and a outline of the surface figure separated at the separating, respectively.

Abstract: Provided are an apparatus and method for more accurately extracting edge portions from an image in various conditions. The apparatus includes: a brightness calculation unit calculating a representative brightness value of the neighborhood of a current pixel; a threshold value calculation unit calculating a threshold value, based on which it is determined whether the current pixel is an edge, by using the calculated representative brightness value; a mask application unit calculating a masking value by applying a mask to an area containing at least the current pixel; and an edge determination unit determining whether the current pixel is an edge by comparing the masking value and the threshold value.

Abstract: A scanning electron microscope comprises an image processing system for carrying out a pattern matching between a first image and a second image. The image processing system comprises: a paint-divided image generator for generating a paint divided image based on the first image; a gravity point distribution image generator for carrying out a smoothing process of the paint divided image and generating a gravity point distribution image; an edge line segment group generation unit for generating a group of edge line segments based on the second image; a matching score calculation unit for calculating a matching score based on the gravity point distribution image and the group of edge line segments; and a maximum score position detection unit for detecting a position where the matching score becomes the maximum.

Abstract: Processing method of a digital image to filter red and/or golden eye artifacts, the digital image comprising a plurality of pixel each comprising at least one digital value represented on a plurality of bits, the method comprising: a step of selecting at least one patch of pixels of the digital image comprising pixels potentially representative of a red and/or golden eye artifact; a step of classifying the at least one patch of pixels as “eye” or “non-eye”; a step of filtering said potentially representative pixels if said patch of pixels is classified as “eye”; wherein the classifying step comprises the operations of: converting the digital values of said patch of pixels into a Gray Code representation, overall obtaining a plurality of bit maps from said patch of pixels, each bit map being associated with a respective bit of said Gray Code; an operation of individually comparing said bit maps with corresponding bit map models belonging to a patch classifier produced by a statistical analysis of bit maps obta

Abstract: A problem of degradation in the accuracy of video matching, which is caused when videos contain video patterns commonly appearing in various videos or video patterns in which features cannot be acquired stably, is solved. In order to solve this problem, a visual feature extraction unit extracts a visual feature to be used for identification of a video based on features of a plurality of pairs of sub-regions in the video, and a confidence value calculation unit calculates a confidence value of the visual feature based on the features of the plurality of pairs of sub-regions. When matching is performed, visual features are compared with each other in consideration of the confidence value.

Abstract: A surface model may be created from aerial photographs, while mitigating the staircase effect, by interpolating sub-pixel values in the photographs and using those sub-pixel values to calculate the offset between overlapping photographs. Aerial photographs are taken, which include overlapping regions. For a pair of photographs, the photographs are rectified to create coordinate systems in which the photographs have x-axes that coincide and y-axes that are parallel to each other, so that overlapping regions in one photograph are a fixed offset along the x-axis from the other photograph. Analytic functions are used to interpolate values between pixels, and the offset distance is calculated by finding the offset that maximizes a similarity function over the analytically interpolated values. The calculated offset may then be used to calculate the height of a point on the photographed surface. A surface model may be built by calculating the heights of many points.

Abstract: A system and method of reducing noise are disclosed. In one embodiment a system comprises an input configured to receive an input image, a filter configured to filter the input image to generate a filtered image, a weight generator configured to generate one or more edge weights related to at least one edge of the input image, a mixer configured to generate an output image based on the input image, the filtered image, and at least one of the edge weights, and an output configured to output the output image.

Abstract: A method for automatically retrieving interaction information between objects, including: with a server, transforming a first image and a second image submitted to said server from a source into first and second sets of parameters, respectively; searching a database for an interaction relationship between the first and second images using the first and second sets of parameters; and returning a representation of the interaction relationship to the source.

Abstract: The present invention provides a method for detecting a specific object in an image to be detected, including: a feature extraction step for extracting an image feature of the image to be detected; and a detection step for detecting detection windows with various sizes of the image to be detected according to the extracted image feature by using classifiers with various sizes corresponding to at least a part of the detection windows with various sizes, so as to determine the presence and location of a specific object in the image to be detected. The invention further provides an object detection device and a system including the device. The method, device and system for detecting a specific object in an image to be detected can improve the precision and increase the speed of the object detection.

Abstract: A ridge flow based fingerprint image quality determination can be achieved independent of image resolution, can be processed in real-time and includes segmentation, such as fingertip segmentation, therefore providing image quality assessment for individual fingertips within a four finger flat, dual thumb, or whole hand image. A fingerprint quality module receives from one or more scan devices ridge-flow—containing imagery which can then be assessed for one or more of quality, handedness, historical information analysis and the assignment of bounding boxes.

Abstract: A system for inspecting a product that includes a scanner to scan across the product and generate an image of the product, the image including a matrix of pixels each having a grayscale value, and an analyzing device electrically coupled with the scanner to receive and analyze the image of the product, the analyzing device including a microprocessor to calculate a reference grayscale value associated with each of the pixels in the image, wherein the reference grayscale value includes an average of the grayscale values of pixels located adjacent to the each pixel, a memory to store the reference grayscale value associated with each of the pixels and at least one threshold associated with the reference grayscale value, and a comparing module to compare the grayscale value of each of the pixels in the image with the at least one threshold associated with the each pixel.

Abstract: Systems and methods are provided for creating contour images that represent the contour of objects reflected in images, calculating contour histogram descriptors of the contour images, and classifying images based in part on the histogram descriptors of the contour images. For example, a contour image of an image is created. A radial-polar grid having a plurality of radial-polar bins is then positioned on the contour image. A contour histogram descriptor is created to include a number of bins that correspond to the radial-polar bins of the radial-polar grid, where the contents of the bins of the contour histogram descriptor represent the number of pixels of the contour image that are located in the corresponding radial-polar bins of the radial-polar grid. Images are classified at least based in part on comparisons between contour histogram descriptors of the images and contour histogram descriptors of training images.

Abstract: Among other disclosed subject matter, a computer-implemented method for pattern matching includes receiving a pattern image, a mask image and a search image, the mask image having an arbitrary shape and identifying a portion of the pattern image. The method includes evaluating a normalized cross-correlation equation based on the pattern image, the mask image and the search image, including at least a convolution of the mask image and the search image. The method includes outputting a result of evaluating the normalized cross-correlation equation, the result indicating whether the search image matches the portion of the pattern image.