Abstract: An approach is provided for characterizing discontinuities in semiconductor devices, for example in a metal silicide. An image of an integrated circuit is caused, at least in part, to be received. The image is analyzed for at least one discontinuity in the integrated circuit structure. A relative measure of the at least one discontinuity is determined in comparison to the integrated circuit structure based on analyzing the image.

Abstract: An image processing apparatus is disclosed which includes: an acquisition section configured to acquire image data of an image; a histogram calculation section configured to calculate a histogram of the acquired image data; a histogram storage section configured to store the calculated histogram that is updated sequentially; and a change determination section configured to determine whether a change has occurred in the acquired image based on the degree of similarity between the calculated histogram and the previously stored histogram.

Abstract: An information processing apparatus and method for processing an input image by comparing a stored background image with the input image to obtain a difference, collating a change in the difference and time of the change with conditions for the occurrence of a periodic event and information as to the time of occurrence of the periodic event, which are previously stored in a storage unit, to determine whether the change in state of the input image is the periodic or a non-periodic movement of an object, and selecting, according to a result of the determination, any one of operations according to the periodic or the non-periodic movement of an object.

Abstract: A method of matching fingerprints is disclosed. The method comprises, for a first minutia point, being assigned as a planet minutia point a) determining pairs including the planet minutia point and a satellite minutia point, respectively, such that a cluster is formed; b) comparing the clusters of the respective sets and excluding non-matching satellites; c) counting links in the cluster formed by remaining pairs; and d) for remaining satellite minutia points, performing steps a) to c) with respective satellite minutia point assigned as planet minutia point to form a supercluster by iterating steps a) to d) and superadding the clusters; calculating a score of the supercluster based on the aggregate counted links; and comparing the score with a threshold. A biometric matching apparatus, a portable data carrier, a data processing unit comprising a matching apparatus and a computer program for implementing the invention are also disclosed.

Abstract: Methods, systems and computer program products are described for providing a computer implemented control system for an automated system. The system includes a document receiving component configured for receiving a plurality of design documents and for rendering a plurality of design document views corresponding to the plurality of design documents. The system also includes an input receiving component configured for receiving an indication selecting first and second design document views, a data extraction component configured for automatically extracting design data from the first and second design document views, a matching component configured for determining a relationship between the design data from the first and second design document views, and a data manager component configured for storing the relationship in a data store accessible by a computer implemented control system that controls a plurality of devices in the automated system based on the stored relationship.

Abstract: The present invention provide an automatic, frame-by-frame, rapid, and accurate search method for video clip. It comprises the following steps. First, acquire and store a plurality of videos to a storage unit. Each video comprises a plurality of frames, respectively. Then, a processing unit extracts the characteristics of the plurality of frames and gives a plurality of characteristic values. Next, the processing unit calculates a plurality of difference values between the characteristic value of each frame and that of the previous frame and between the characteristic value of the subsequent frame and that of the current frame. Store a data set, which is produced by corresponding the plurality of difference values between all frames of each video to the time axis, to the storage unit. Finally, the processing unit compares a data set of any video with a plurality of data sets of other videos and gives a search result.

Abstract: Systems and related methods for detecting a scene in video content are provided. In one example, a scene detection program identifies a plurality of shots in the video content and selects a target shot. The program builds a forward window including the target shot and having shots that are temporally ahead of the target shot, and a rearward window having shots that are temporally behind the target shot. For each of the shots in the forward window the program determines a dissimilarity between a selected shot in forward window and each of the other shots in the rearward window. If one or more of the dissimilarities is less than a scene boundary threshold, the program determines that the scene does not begin at the target shot. If none of the dissimilarities is less than the scene boundary threshold, the program determines that the scene begins at the target shot.

Abstract: An image filtering method and apparatus that filter a block of a plane area by using a pseudo-random filter are provided, whereby artifacts generated in image encoding and decoding procedures are removed, the method including: determining a similarity between pixel values of a the first block; and generating a second block by selectively applying a pseudo-random filter on the first block according to a result of based on the determining the determined similarity.

Abstract: Interest points are markers anchored to a specific position in a digital image of an object. They are mathematically extracted in such a way that, in another image of the object, they will appear in the same position on the object, even though the object may be presented at a different position in the image, a different orientation, a different distance or under different lighting conditions. The goal is to match interest points in one image with corresponding interest points in another image. Typically, this involves the construction of a descriptor, which is both computationally expensive and resource-intensive. Methods and devices are described that match interest points without the construction of conventional descriptors and that permit the use of spatial coherency information to increase the accuracy of the match.

Abstract: In one embodiment, there is provided a method for an Optical Character Recognition (OCR) system. The method comprises: recognizing an input character based on a plurality of classifiers, wherein each classifier generates an output by comparing the input character with a plurality of trained patterns; grouping the plurality of classifiers based on a classifier grouping criterion; and combining the output of each of the plurality of classifiers based on the grouping.

Abstract: An image processing apparatus includes an acquiring unit, a specifying unit, a search unit and a difference extracting unit. The acquiring unit acquires a first image and a second image. The specifying unit specifies one or more image areas included in the first image. The search unit searches the second image for an image area corresponding to each of the one or more image areas specified by the specifying unit. The difference extracting unit extracts a difference between the corresponding image area obtained by the search unit and each of the one or more image areas specified by the specifying unit.

Abstract: A method of image identification based on artificial intelligence is provided, the method includes initializing a second computer with a first computer supplying and accepting image data from the first computer to the second computer; comparing the delivered image data in the second computer to pre-existing stored images in the second computer; generating and displaying a first result if the delivered image data is the same as the stored image in the second computer and a second result if the delivered image data is not the same as the stored image data; and un initializing the second computer based on the first and second results.

Abstract: Bicycles present before a vehicle are identified. A picked-up image having intensity values that are in accordance with the temperatures of objects is acquired to extract image regions having intensity values representative of higher temperatures than a background. An object, which includes a pair of first object portions (legs) and a second object portion (a tire), would be identified as a bicycle if the pair of first object portions, which have greater dimensions in length in the vertical direction than in width in the horizontal direction and the distance between which is less than a predetermined value in the horizontal direction, are detected from the extracted image regions and further if the second object portion, which has a different intensity value from any one of the first object portions and also has a length greater than a predetermined value in the vertical direction, is existent between the pair of first object portions.

Abstract: An image quality adjustment capability measurement method includes the steps of: using an instructor image serving as a target for the image qualities of an evaluation image and a correlation function between image quality characteristic differences and image quality adjustment capability values, the instructor image and the correlation function being obtained in advance; inputting a corrected image obtained after the image qualities of the evaluation image is adjusted; calculating the image quality characteristic difference between the input corrected image and the instructor image; and calculating the image quality adjustment capability value from the correlation function between the image quality characteristic differences and the image quality adjustment capability values, which is obtained in advance, and the calculated image quality characteristic difference.

Abstract: An image recognition apparatus includes: a storage unit configured to store a dictionary used to recognize a predetermined pattern; a recognition unit configured to detect in image data using the dictionary, as recognition results, a plurality of partial regions having a likelihood of being the predetermined pattern greater than a predetermined threshold; a display unit configured to display the image data, and also display information indicating the partial regions that were detected by the recognition unit; a determination unit configured to determine, based on an instruction by a user, a negative region from among the plurality of partial regions that were detected by the recognition unit; a generation unit configured to generate a learning image based on the determined negative region; and an update unit configured to update the dictionary based on the learning image that was generated by the generation unit.

Abstract: Various embodiments provide approaches for measuring the rendering times of a webpage that can be browser-independent. In at least some embodiments, as a web page is rendered, video data of the webpage is captured and analyzed to ascertain when the webpage, or portions thereof, have likely been completely rendered. In at least some embodiments, image processing can be performed on captured video data to ascertain, from observed pixel changes over time, when the webpage has likely been rendered.

Abstract: A method of measuring load time for a web page includes video recording a duration of the web page load including capturing N frames in a series of frames 1 through N of the web page. The method includes comparing each frame from frame 1 through frame N?1 with a next frame from frame 2 through frame N in the series of frames to determine a pixel change amount between the compared frames. The method includes defining a series of page phases for the duration of the web page load, each page phase defined by a start time and an end time with each phase start time determined based on estimating user perceptible pixel change amount levels.

Abstract: Embodiments of the invention are directed to methods and apparatuses for capturing a real-time video stream using a mobile device, determining, using a processor, which images from the real-time video stream are associated with individuals meeting a user defined criteria, and presenting on a display of the real-time video stream, one or more indicators, each indicator being associated with an image determined to be a person meeting the predefined criteria.

Abstract: A system and method for comparing captured sequences of in-vivo images with (e.g., template or model) sequences, for example, for computer-automated recognition of contractions. The size of the opening of an in-vivo lumen passageway represented in each frame in a subset of frames of an image stream captured in vivo may be measured. Frames in the subset of frames of the image stream having a local minimum size of the lumen passageway may be identified. The subset of frames may be divided into segments of sequential frames at frames having local maximum lumen sizes before and after the identified frame having a local minimum size of the lumen passageway to generate contraction sequences. A plurality of the contraction sequences may be compared to template sequences. A plurality of the contraction sequences may be displayed.

Abstract: A system identifies an image and determines whether the image contains inappropriate content based on first data associated with the image, second data associated with a document that contains the image or refers to the image, and/or third data associated with a group of documents with which the image is associated.

Abstract: Methods for processing a video stream are provided. One embodiment comprises generating a comparison signature for a frame of a video stream and comparing the comparison signature to a reference signature. In another embodiment, a method comprises generating a reference signature for a frame of a video stream and assigning an action to the frame containing the reference signature. Example system embodiments for implementing the aforementioned methods are also provided.

Abstract: A facial recognition method for eliminating the effect of noise blur and environmental variations is provided, applicable to a data processing apparatus, to determine whether a current-face-image matches a reference-face-image or not. According to the method, Gaussian Noise blur Reduction or another noise blur reduction method is performed on the current-face-image and the reference-face-image. Furthermore, the current-face-image and the reference-face-image are respectively divided into a plurality of blocks, so as to derive feature-vector sets representing the current-face-image and the reference-face-image. Finally a dynamic threshold value is determined according to environmental variations to be compared with the difference of the feature-vector sets, so as to determine whether a current-face-image matches a reference-face-image.

Abstract: A method for detecting authorized security markers includes capturing an image of a region of interest on a product with a camera; storing image data in a two-dimensional array on a microprocessor; counting a number of pixels at or above a predetermined brightness level in the image data with the microprocessor to determine a first score; eroding the image data; counting the pixels which remain at or above the predetermined brightness level after erosion to determine a second score; calculating a ratio of the second score to the first score; and producing a first authentication signal if the ratio meets a first predetermined criteria.

Abstract: Disclosed is a method (101), in relation to a current video frame (300) comprising a visual element (320) associated with a location in a scene captured in the frame (300), said visual element (320) being associated with a plurality of mode models (350), said method matching (140) one of said plurality of mode models (350) to the visual element (320), said method comprising, for each said mode model (350), the steps of determining (420) a visual support value depending upon visual similarity between the visual element (320) and the mode model (350), determining (440) a spatial support value depending upon similarity of temporal characteristics of the mode models (350) associated with the visual element and mode models (385) of one or more other visual elements (331); and identifying (450) a matching one of said plurality of mode models (350) depending upon the visual support value and the spatial support value.

Abstract: One embodiment of a system of present invention includes means for obtaining an image from a requester, means for obtaining contact information for vouching people, whom requester identifies as having personal knowledge of the requester, and means for sending an electronic message to the vouching people, with a link to an image lineup. The image lineup contains the image obtained from the requester and other images. The system further includes means for receiving a selection choice from the vouching people, identifying whether the image in the image lineup depicts the requester, and means for calculating a statistical rating of a likelihood that the image obtained from the requester depicts him/her.

Abstract: Methods and apparatus for procedural directional texture generation. A procedural directional texture generation method may, for example, be used to design hair or hairstyles. The method may obtain one or more strokes, one or more optional masks, and one or more optional user maps. One or more interpolated orientation maps may be generated from the input. The orientation maps, possibly along with one or more optional user maps, may be used to compute synthetic low-frequency lighting. A noise map may be generated at one or more frequencies and used, along with the interpolated maps, to generate high-frequency lighting. Alternatively, a flow-guided texture synthesis method may be used to generate high-frequency lighting. The low- and high-frequency lighting may then be combined to generate a desired frequency spectrum. Color may be added to the full structure, or alternatively color may be added at an earlier step.

Abstract: An image retrieval system includes an image entry unit configured to input an image. A descriptor determining unit receives the image and transforms the image into an intermediate image to generate a descriptor of the image, for example, utilizing the distance transform (DT) application. An image comparing unit compares the image with a plurality of candidate images by comparing the descriptor of the image with descriptors of the plurality of candidate images. Finally, a retrieving unit retrieves one or more candidate images similar to the input image based on the comparison of their descriptors.

Abstract: A system and method for generating and using a correlation filter. The method includes providing a plurality of training images, each training image being paired with an associated target correlation plane. Each training image and target correlation plane pair is processed. A final filter is generated, wherein the final filter is useable to generate a filtered output correlation plane of each training image. The final filter is selected to collectively minimize errors between the filtered output correlation plane of each training image and its associated target correlation plane. The final filter can be used in a wide variety of still image and video based object location and tracking applications.

Abstract: Cell phones and other portable devices are equipped with a variety of technologies by which existing functionality can be improved, and new functionality can be provided. Some relate to visual search capabilities, and determining appropriate actions responsive to different image inputs. Others relate to processing of image data. Still others concern metadata generation, processing, and representation. Yet others relate to coping with fixed focus limitations of cell phone cameras, e.g., in reading digital watermark data. Still others concern user interface improvements. A great number of other features and arrangements are also detailed.

Abstract: One embodiment of the present invention envisions providing requester with a computer program for a remote computer equipped with a scanning device. The computer program controls the scanning device and prevents alterations to an image produced by the scanning device. The unaltered image of a photo ID is transmitted from the remote computer to the verifier's server computer. The image of the photo ID may further be used to verify the requester by comparing the image to photographs from other sources.

Abstract: A hierarchical motion deblurring method for a single image is provided. In the method, a blur kernel of a target image is calculated and a multi-scale representation for representing the target image and the blur kernel is constructed. Then, a gradient attenuation function and a strong edge suppression function are applied to a residual Richardson-Lucy algorithm, so as to iteratively calculate a residual image between the blur kernel and the target image represented by the representation in each scale and restore the residual image to obtain a first restored image and a second restored image. Finally, the two restored images are compared so as to obtain a motion deblurring image.

Abstract: A method for emphasizing differences in graphical appearances between an original document and a modified document using annotations with outer and inner boundaries is disclosed. A first bitmap of the modified document and a second bitmap of the original document are received. A set of difference points is derived from a comparison of the first bitmap and the second bitmap. Those difference points are representative of coordinates within the first and second bitmaps having pixel value differences exceeding a predetermined threshold. An outer boundary and an inner boundary of the difference points is derived, then an annotation defined by an area between the outer boundary and the inner boundary is generated.

Abstract: Techniques for capturing images of business cards, uploading the images to a designated computing device for processing and recognition are disclosed. A mechanism is provided to update extracted data from the images when there are any changes. Depending on implementation, there are a number of ways to capture images of business cards (e.g., via a phone camera, a PC camera, or a scanning device). A transmission means is provided to transport the images to the designated computing device for centralized management of integrated contact information for individual users. As a result, a user may access his/her updatable integrated contact information database anywhere anytime from a chosen device.

Abstract: Approaches for enabling a computerized entity to recognize characters in an electronic document. In a persistent data store, character identification data is stored. Character identification data is data that, for one or more characters of one or more fonts, associates (a) glyph data for a character with (b) code point data for the character, where the glyph data describes how to render the character on or to an output device, and the code point data identifies, to the computerized entity, the identity of the character. Upon determining that an embedded font document, such as a PDF document, does not include a set of code point data for a particular character, the character identification data is consulted to determine the identity of the particular character. In this way, a machine can recognize characters in the embedded font document and perform functions such as indexing or searching on the embedded font document.

Abstract: A method of automatically identifying a scan region boundary, used for scanning a document under a non-unified background, includes: reading a reference linear scan data by a scanning device, the scanning device analyzing brightness values in each row of the reference linear scan data and determining a reference range value according to the brightness values in each row; and during scanning a document, the scanning device comparing brightness values of a linear scan data with the reference range value so as to determine the scan region boundary. The method confirms the scan boundary of a document and is not subjected to a backlight element or a document carrier. The method identifies the boundary of a document under a non-unified background, is applicable to scanning transparent and reflective documents and has strong resistance against environmental interference, not only simplifying the scanning device but also enhancing the automatic performance.

Abstract: Provided is a method for determining the magnification of a pattern searching template of a scanning electron microscope. The determining method comprises: acquiring a first image initially at a first magnification; then acquiring a second image which contains a pattern image displayed on the first image at a second magnification lower than the first magnification; making the size of the first image coincident with the size of a third image which cut out a portion of the second image; thereafter determining the correlation value between the first image and the third image; and setting the second magnification as the magnification of a pattern searching template, in the case where the correlation value is equal to or higher than a predetermined value. As a result, a condition for acquiring a search area can be properly set, when pattern recognition is performed by means of the template.

Abstract: Method for measuring the dissimilarity between a first and a second images, including the following steps: a) multiresolution decomposition of the first and the second images to obtain coefficients of the first and of the second images, each coefficient being function of a scale and a location in space; b) constitution of the patches for the first and the second images; c) evaluation of the dissimilarity between the probability density functions of patches having a given scale and belonging to the first image and of patches having the same scale and belonging to the second image, the dissimilarity being a partial measure of the dissimilarity between the first and the second images; and a method for measuring the dissimilarity between a first and second video sequences, the method following a similar multi-scale approach based on sparse intrascale/interscale/interchannel patches and additionally taking motion into account.

Abstract: The present invention relates generally to identifying incorrect or fraudulent business listings. More specifically, business listings may be validated by comparing the listing information to information identified from images at or near the location of the business. For example, a plurality of geolocated street level images of various locations may be categorized using computer vision techniques. A particular business listing may include, among other information, a location (address, etc.) as well as a business category. The pre-categorized images are then selected based on the location of the particular business listing. The categories of the selected images are then compared to the category of the particular business listing to validate the business listing.

Abstract: An ultrasound imaging system and method provides for extended imaging.

Abstract: According to one embodiment, a pattern recognition method includes calculating similarities of the input pattern with respect to respective categories, converting the calculated similarities of the input pattern with respect to the respective categories into first evaluation values based on a first table which indicates a relationship between similarities for respective categories and first evaluation values, calculating second evaluation values based on the calculated first evaluation values for the respective categories and prior probabilities for the respective categories stored in a second table indicating prior probabilities of the respective categories, and selecting a category corresponding to a maximum value of the calculated second evaluation values.

Abstract: Various embodiments provide approaches for measuring the rendering times of a webpage that can be browser-independent. In at least some embodiments, as a web page is rendered, video data of the webpage is captured and analyzed to ascertain when the webpage, or portions thereof, have likely been completely rendered. In at least some embodiments, image processing can be performed on captured video data to ascertain, from observed pixel changes over time, when the webpage has likely been rendered.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for labeling images. In one aspect, a method includes automatically identifying an object in an image using a deep model-based and data-driven hybrid architecture.

Abstract: A method of identifying potential phishing abuse images includes: producing a first color map that represents a subset of color values and pixel locations within a base image; producing a second color map that represents color values and pixel locations within a target image; selecting an alignment the first color map with the second color map such that at least some pixel locations of the first color map align with at least some pixel locations of the second color map; determining a measure of color value matching of aligned pixel locations for the selected alignment; and repeating the acts of selecting and determining until a prescribed threshold measure of color value matching is determined for at least one of the selected alignments or until an evaluation limit is reached.

Abstract: To generate at least one hash value for a feature vector that represents a data object, a discrete orthogonal transform is applied on a second vector produced from the feature vector. Applying the discrete orthogonal transform on the second vector produces a third vector. At least one value is selected from the third vector to produce the hash value. The at least one hash value is used to perform an action.

Abstract: Techniques are disclosed for detecting new events in a video stream that yield improved detection efficiency in real time. For example, a method determines whether a given event is a new event in a video stream. The video stream includes a plurality of events. A first step extracts a first set of features (e.g., text features) from the given event. The first set of features is computationally less expensive to process as compared to a second set of features (e.g., image features) associated with the given event. A second step computes one or more first dissimilarity values between the given event and one or more previous events in the video stream using only the first set of features when one or more first dissimilarity criteria exist. A third step determines whether the given event is a new event based on the one or more computed first dissimilarity values.

Abstract: A method comprises searching in a video stream a first frame and a second frame that each has enough point correspondence with a image model, wherein the first frame is the nearest previous frame prior to a third frame, and the second frame is the nearest subsequent frame to follow the third frame. The method further comprises calculating an interpolation between a first mapping matrix of the first frame and a second mapping matrix of the second frame to obtain a third mapping matrix of the third frame that has insufficient point correspondence with the image model.

Abstract: A method for providing guidance to a user as to what images should be captured by the user as the user moves from location to location, comprising receiving a first captured digital image having pixel data captured of a scene at a first location, using a processor to analyze the pixel data of the first captured digital image and determining based on such analysis a second possible image capture location for the user, wherein the second possible image capture location is selected from a plurality of different possible locations, and transmitting guidance to the user over a communications network, the guidance including information concerning a next scene to be captured at the second location.

Abstract: Detecting a static graphic object (such as a logo, title, or sub-title) in a sequence of video frames may be accomplished by analyzing each selected one of a plurality of pixels in a video frame of the sequence of video frames. Basic conditions for the selected pixel may be tested to determine whether the selected pixel is a static pixel. When the selected pixel is a static pixel, a static similarity measure and a forward motion similarity measure may be determined for the selected pixel. A temporal score for the selected pixel may be determined based at least in part on the similarity measures. Finally, a static graphic object decision for the selected pixel may be made based at least in part on the temporal score.

Abstract: An image processing apparatus includes: a category area dividing unit that divides each image constituting the time-series images of an object to be examined, which are captured in time series, into category areas including an examined area and an abnormal candidate area, the examined area showing the object to be examined, the abnormal candidate area being possibly an abnormal area showing an abnormal site of the object to be examined; an optical flow calculator that calculates optical flows of the examined area and the abnormal candidate area from images constituting the time-series images; and an abnormal area detector that compares the optical flow of the examined area and the optical flow of the abnormal candidate area, and detects the abnormal candidate area as the abnormal area according to the result of the comparison.

Abstract: An image search apparatus provides searching for a search-target image corresponding to an input image from among a plurality of search-target images. The image search apparatus includes a characteristic partial image detection unit and a search unit. The characteristic partial image detection unit detects a characteristic partial image of each search-target image based on a dissimilarity level of a partial image at a corresponding position among a plurality of search-target images. The search unit respectively calculates a level of coincidence between a characteristic partial image of each search-target image detected by the characteristic partial image detection unit and a partial image of an input image. The search unit further searches for a search-target image corresponding to an input image from among a plurality of search-target images based on the coincidence level.