Abstract: The present invention relates to methods and devices for analyzing x-ray images. In particular, devices, methods and algorithms are provided that allow for the accurate and reliable evaluation of bone structure from x-ray images.

Abstract: A coin detecting apparatus has: a light source for emitting a light beam onto either of two surfaces of a coin; a half mirror for splitting the light beam reflected on the surface of the coin into a transmitted light beam and a reflected light beam; a full area sensor for obtaining a full image of the surface of the coin in accordance with one of the transmitted light beam and the reflected light beam; and a regional area sensor for obtaining a partial image of the surface of the coin in accordance with the other one of the transmitted light beam and the reflected light beam. By doing this, it is possible to provide a coin detecting apparatus capable of improving accuracy in coin detection and shortening time therein.

Abstract: A thermal imaging system for a battery module enclosure that includes first and second battery module enclosure components between which a weld is formed includes a thermal imaging camera that focuses on the first and second battery module enclosure components within a predetermined amount of time after the weld is formed and that acquires a thermal signature. A control module includes an image processing module that receives the thermal signature and that locates a predetermined reference point in the thermal signature. An image comparison module receives the thermal signature and uses the predetermined reference point to compare the thermal signature to a template signature in order to verify structural integrity of the weld. The image comparison module computes a relative measure of deviation of the thermal signature from the template signature and identifies the weld as defective when the relative measure of deviation is greater than a predetermined value.

Abstract: An apparatus for automatically assessing the quality of a printed circuit board assembly using digitized video image analysis. The apparatus integrates with existing relatively low precision automated surface mount technology (“SMT”) manufacturing systems as an inspection station insertable at various steps in the assembly process or as a separate manually loaded station. The inspection station includes a high resolution video imaging system and a video image analyzer comprising an onboard master computer that generates control signals to reposition the camera mounted within a screen on a movable carriage and/or reposition the circuit board, and adjust the lighting; and generates individual board status data to be archived, graphically displayed on monitors or otherwise utilized by a rework station.

Abstract: It is an object of the present invention to provide a method and apparatus for performing pattern inspection which reduces the number of pseudo defects. A pattern inspection apparatus according to an embodiment of the present invention comprises an optical image acquiring unit which acquires optical image data of a target plate to be inspected, the target plate being formed as a pattern, a design image data generating unit which generates first design image data based on a first design pattern serving as a base for pattern formation of the target plate, and a comparing unit which compares the optical image data with the first design image data. In the comparing unit, second design image data generated based on a second design pattern is further input, and the optical image data is compared with the second design image data in place of the first design image data.

Abstract: It is an object of the present invention to provide a method and apparatus for performing pattern inspection at an appropriate precision. A pattern inspection apparatus according to an embodiment of the present invention comprises an optical image acquiring unit which acquires optical image data of a target plate to be inspected, the target plate being formed as a pattern, a design image data generating unit which generates design image data based on a design pattern serving as a base for pattern formation of the target plate, and a comparing unit which compares the optical image data with the design image data. In the comparing unit, region image data generated based on the information of the region pattern representing a predetermined region and formed in the same format as that of information of the design pattern is input, and a determination condition is changed based on the region image data when the optical image data is compared with the design image data.

Abstract: The present invention relates to a defect inspection apparatus for inspecting defects in patterns formed on a semiconductor device, on the GUI of which for the confirmation of the inspection results an area is provided for displaying any one of or facing each other the features amount of defects, and the image during inspection or the reacquired image, and on the GUI of which a means is provided for setting the classification class and importance of the defects, and based on the classification class and the importance of the defects information set by this setting means, the classification conditions or the defect judging conditions are automatically or manually set so that the inspection conditions may be set easily.

Abstract: A defect detecting method includes: an image acquiring process for capturing an image of an inspection object having an identical sequence of pattern and acquiring the image; and a defect emphasizing process for emphasizing a defect of the captured image. The defect emphasizing process includes: an inspected-point selecting process for sequentially selecting an inspection point on the captured image; and a defect emphasizing value calculating process for obtaining differences by subtracting from a luminance value of the selected inspection point each luminance value of a plurality of comparison points disposed around the inspection point and selecting one difference of the smallest value of the obtained differences in luminance to determine a defect emphasizing value of the inspection point.

Abstract: A convolutional neural network is implemented on a graphics processing unit. The network is then trained through a series of forward and backward passes, with convolutional kernels and bias matrices modified on each backward pass according to a gradient of an error function. The implementation takes advantage of parallel processing capabilities of pixel shader units on a GPU, and utilizes a set of start-to-finish formulas to program the computations on the pixel shaders. Input and output to the program is done through textures, and a multi-pass summation process is used when sums are needed across pixel shader unit registers.

Abstract: According to one aspect of the present invention, there is provided a digital camera system being capable of taking a frame of image data comprising a plurality of color elements, the system being arranged: to divide the frame image data into a plurality of blocks including a plurality of pixel data; to calculate a predetermined value for the each color element and for each of all or a part of the plurality of blocks; to judge, for each of all or a part of the plurality of blocks, whether the block being likely a part of a grey object or not to cumulate the predetermined values of the blocks judged as being likely a part of a gray object for the respective color elements, and; to decide a first set of digital gains for adjusting white balance based on the cumulated values.

Abstract: An image processing apparatus comprises a dye amount distribution obtaining section configured to obtain a distribution of dye amounts from a sample colored with a dye. The apparatus further comprises a classification section configured to classify a point on the sample in accordance with a dye amount at a corresponding position on the distribution obtained by the dye amount distribution obtaining section.

Abstract: The image processing apparatus according to the present invention includes: a segmentation process section for performing a segmentation process with respect to image data corresponding to an image to be processed; first and second input tone correction sections each of which performs a density correction process with respect to the image data by referring to a conversion curve which indicates a relation between density before correction and density after correction; and a processing mode control section for switching between (i) a first processing mode in which the segmentation process is performed with respect to the image data having been subjected to the density correction process and (ii) a second processing mode in which the segmentation process is performed with respect to image data having not been subjected to the density correction process. Therefore, the image processing apparatus can perform the segmentation process with high accuracy.

Abstract: A method of evaluating element distribution within a data set, such as pixel characteristics whose scatterness is quantified. Scatterness is evaluated for individual scale levels as the image is merged through smaller scale levels. At each level, every pixel block (i.e., 2×2) in the region of interest merges to one pixel to form the next level. Select characteristics of the pixels are assessed in the merge, and the image may be simplified to fewer pixel states so the merges require less processor overhead. In one merge mode if any pixel in the block has the desired characteristic, then the resulting merged pixel is set with that characteristic. The merging rate of the pixels is used to determine the scatterness for the pixel distribution. A scatterness sequence can be used separately or in combination with known techniques for comparing images in applications such as search and classification.

Abstract: An image processor for generating a histogram at high speeds under a low processing load. The image processor is used with a memory and generates sub-sampling image data and histogram data based on captured image data. The captured image data includes gradation values of a plurality of pixels. The image processor includes a histogram data acquisition unit connected to the memory. The histogram data acquisition unit detects pixels having the same gradation value from the captured image data, designates an address in the memory that corresponds to the gradation value of the detected pixels, and adds or subtracts an arbitrary value to or from a value stored at the designated address of the memory.

Abstract: An image display device capable of supporting brightness enhancement and power control and a method thereof, in which the brightness and contrast are enhanced and the power of a light source is controlled in accordance with the image characteristic. In the image display method, a parameter representative of an input image is calculated from a histogram of the input image, a model corresponding to the input image is selected from a plurality of representative models by analyzing the parameter, and then a brightness controlling function matched with the selected model is selected. The size of each component of a pixel constituting the input image is controlled using the brightness controlling function, and the output image of the pixel whose component size has been controlled is displayed.

Abstract: A color information creating part calculates the average color of captured image data in units of macro blocks, and determines to which color index the average color of each macro block corresponds. Color arrangement information formed by associating color information with the macro blocks is stored in a color index memory. A histogram creating part creates color histograms representative of the distribution of macro blocks added with color information corresponding to aimed color indexes. Based on the color arrangement information and the color histograms, a color application execution part recognizes objects, surroundings, or the situation of a driver's own vehicle without performing area splitting (shape acquisition) of an image.

Abstract: A method for segmenting a digitized image includes providing a digitized image comprising a plurality of intensities corresponding to a domain of points on an N-dimensional grid, defining a weight between each pair of adjacent points in the image based on the intensity difference of each said point pair, marking a subset of points in said image wherein each marked point is marked with one of a plurality of labels, determining the probability that an unmarked point should be marked with one of said plurality of labels using a multigrid method that uses a restriction operator to map said provided image to a coarse-level gird, and a prolongation operator to map said coarse-level image back to a fine-level grid, wherein said restriction operator is an adjoint of said prolongation operator, and wherein a fine-level segmentation is a prolongation of a coarse level segmentation.

Abstract: A background generating method of low process cost for use in detecting a moving object based on subtraction process performed between an input image and a base image, and an apparatus with the function of detecting a moving object using its background. The apparatus consists of a feature vector extractor for calculating feature vectors from the input image and an average processor capable of obtaining an average from the calculated feature vectors, diminishing noise occurring for a short time and the influence by the moving object, and forming images that follows the illumination change due to changes in the imaging environment. The images generated by the average processor are accumulated for a predetermined number of frames and the accumulated images are statistically processed by a statistical processor.

Abstract: An apparatus, system, and method are disclosed for identifying content within a scanned document. The apparatus includes a modification module, an identification module, and a segmentation module. The modification module creates a modified content data set through application of a sigmoid function to a scanned content data set. The identification module identifies a content segment within the modified content data set. The segmentation module identifies a content segment type of the content segment. Exemplary content segment types include text, line art, and images.

Abstract: A method for classifying regions defined within a digital image includes defining a tessellated set of regions from a plurality of regions identified by a plurality of segmented versions of the digital image and then classifying each tessellated region according to a likelihood that the tessellated region contains text.

Abstract: A document includes a region with a high character recognition ratio, and that with not so high a character recognition ratio. Hence, even when identical types of information are provided because they have the same attribute, it is hard to say that satisfactory information provision can be made. Also, different kinds of information are required depending on output methods. To solve these problems, output information to be output for each region is determined based on information of an output destination of a document image and attributes of respective regions in the document image, and pieces of output information to be output are combined to be output to the output destination. The output information to be output for each region may be determined based on the information of the output destination of the document image, the attributes of the respective regions in the document image, and user's authority information.

Abstract: An image recognition apparatus, includes: a reading unit that reads an image on a paper script to acquire the image data from the paper script; an additional image extraction unit that extracts an additional image added to each predetermined area of an original image from the difference extraction result by taking a difference between an image data of the original image on the paper script and the image data acquired by the reading unit; an association unit that associates the additional image with the predetermined area; and an additional image selection unit that selects one of a plurality of additional images under prescribed decision conditions, when the plurality of additional images are associated with one predetermined area in accordance with an association made by the association unit.

Abstract: A Mixed Media Reality (MMR) system and associated techniques are disclosed. The MMR system provides mechanisms for forming a mixed media document that includes media of at least two types (e.g., printed paper as a first medium and digital content and/or web link as a second medium). In one embodiment, an MMR document is retrieved based on recognition of a paper document. Responsive to a comparison of the paper document and a virtual multimedia document, a set of actions is displayed to a user. Responsive to a user selection, the user-selected action is performed. In another embodiment, a captured list of names is used to assist a user with labeling a media file.

Abstract: Techniques for processing handwriting input based upon a user's writing style. Some techniques employ the style in which the user writes a single character, while other techniques alternately or additionally employ a group of allographs that form a handwriting style. Some implementations of these techniques, such as those implemented in writing style analysis tool, analyze one or more characters written by a user to identify a community, such as a geographic region or cultural group, to which the user's handwriting style belongs. Other implementations analyze one or more characters of a user's handwriting in order to alternately or additionally categorize the user's handwriting into a particular handwriting style. The writing style analysis tool may then provide the user with a handwriting recognition application specifically configured for that user's personal handwriting style.

Abstract: A Mixed Media Reality (MMR) system and associated techniques are disclosed. The MMR system provides mechanisms for forming a mixed media document that includes media of at least two types (e.g., printed paper as a first medium and digital content and/or web link as a second medium). In one particular embodiment, the MMR system includes a method, system, and computer program product for adding a hotspot to an imaged document. A source document is converted to an imaged document, from which features are extracted. Hotspots are added to the imaged document, and the imaged document, hotspot definitions, and the feature representation are stored.

Abstract: A Mixed Media Reality (MMR) system and associated techniques are disclosed. The MMR system provides mechanisms for forming a mixed media document that includes media of at least two types (e.g., printed paper as a first medium and digital content and/or web link as a second medium). In one particular embodiment, the MMR system includes a content-based retrieval database configured with an index table to represent two-dimensional geometric relationships between objects extracted from a printed document in a way that allows look-up using a text-based index. A ranked set of document, page and location hypotheses can be computed given data from the index table. The techniques effectively transform features detected in an image patch into textual terms (or other searchable features) that represent both the features themselves and the geometric relationship between them. A storage facility can be used to store additional characteristics about each document image patch.

Abstract: A method and system for generating land cover classifications is provided. Embodiments of the invention are directed toward a method in a computing environment for determining land cover classification that includes identifying a geographic unit and determining a human presence datum corresponding to the geographic unit. The method can further include determining an environmental datum corresponding to the geographic unit and/or a cultural datum corresponding to the geographic unit. The method can still further include assigning a land cover classification to the geographic unit based on at least the human presence datum and (a) the environmental datum, (b) the cultural datum, or (c) both (a) and (b). In certain embodiments, the method can still further include determining a geographic region corresponding to the geographic unit and assigning a land cover classification to the geographic unit based on at least the geographic region.

Abstract: According to the image processing apparatus, method and program of the present invention, images to be used can be automatically selected from inputted images for the number of images designated by a user. Moreover, a user can easily manually add an image to be used so that the number of images to be used is equal to or less than the number previously designated by the user.

Abstract: False positive detection of discrimination targets within images is reduced, while detection processes are accelerated. A partial image generating means generates a plurality of partial images by scanning a subwindow over an entire image. A candidate classifier judges whether each of the partial images represent a face (discrimination target), and candidate images that possibly represent faces are detected. A discrimination target discriminating means judges whether each of the candidate images represents a face. The candidate classifier has performed learning, employing reference sample images and in-plane rotated sample images.

Abstract: A method of searching and identifying a signal includes the steps of providing a signal having one or more signal units and identifying at least one of the signal units by a predetermined pattern having a first order. Next, comparing a group of the one or more signal units comprising the at least one signal unit, to one or more predetermined patterns having an order higher than the first order, and if a match is found, modifying the signal by replacing the group of the one or more signal units by a higher order signal unit. Finally, identifying the higher order signal unit by a higher order pattern, wherein the higher order pattern comprises information identifying the group of one or more signal units. The predetermined patterns are ordered according to a hierarchy, wherein a higher order pattern comprises at least one lower order pattern. The signal may be electronic data, transient signals, digital signals, analogue signals, text, spectra, gene sequences or genetic data.

Abstract: An apparatus detects a predetermined number of facial images from detection target images. An inclination order setting means utilizes the correlative relationships among correlative data obtained by a correlative data obtaining means and the inclinations of faces that appear in input images, to determine the relative value of the probability that faces of a predetermined inclination will appear in the input images. The inclination order setting means sets the order of inclinations of faces to be detected such that faces are detected in order of inclinations according to the relative values of the probabilities, based on the correlative data that they will appear. A face detecting means detects faces within the input images while varying the inclinations of faces to be detected according to the set order.

Abstract: An image processing apparatus includes a production unit that can produce data by putting original image data into a first region and copy-forgery-inhibited-pattern image data into a second region, and a processing unit that can perform predetermined processing to the original image data put in the first region. The image processing apparatus further includes a composite image data producing unit that can produce composite image data based on both of the original image data to which the predetermined processing is performed and the copy-forgery-inhibited-pattern image data put in the second region of the data.

Abstract: An image processing apparatus includes a decision unit configured to determine whether or not print image data includes copy-forgery-inhibited-pattern image data. Based on a decision result of the decision unit regarding whether or not the print image data includes the copy-forgery-inhibited-pattern image data, a processing unit of the image processing apparatus is configured to determine whether first compression processing is to be performed or second compression processing is to be performed and apply the decided compression processing to the print image data.

Abstract: A data compression apparatus includes a thinning section that thins numerical values from a succession of values constituting compressed data to create first and second compressed data; a reversible compression section for the first compressed data; a detecting section that detects an edge portion of the image; and a non-reversible compression section that processes the second compressed data based on a result of the edge detection. When the detecting section detects no edge portion, the non-reversible compression section outputs a predetermined code of a number of bits. When the detecting section detects the edge portion, the non-reversible compression section outputs a first value based on image numerical value when involved with a predetermined positional relation expressed by a first number of bits and a second value expressed by a few number of bits than the first number when the image numerical value is not involved with a predetermined positional relation.

Abstract: An image data processing device for scaling up or down input image data and outputting output image data includes a first image processing section for converting the input image data with a first resolution into intermediate image data with a second resolution, and a second image processing section for converting the intermediate image data into the output image data with a third resolution.

Abstract: An image processing apparatus in which image data is transformed to spatial frequencies to obtain multiple frequency components, the obtained frequency components are processed, the processed frequency components are inversely transformed to image data, and the number of gray levels of the inversely transformed image data is decreased, and the image processing apparatus comprising a frequency component judgment section for comparing the absolute value of each of specific multiple frequency components in the multiple frequency components obtained after transformation to spatial frequencies with a predetermined value in terms of magnitude, a frequency component changing section for changing each of specific frequency components on the basis of the result of the comparison, and a noise addition section for adding a specific value to each of the specific frequency components having been changed and other frequency components.

Abstract: An image processing apparatus includes a down-converter and an image converter converting first image data into higher quality second image data. The image converter includes a prediction tap extracting unit extracting, as prediction taps, a plurality of pixels used for predicting a subject pixel of the second image data from the first image data, a class tap extracting unit extracting, as class taps, a plurality of pixels used for classifying the subject pixel from the first image data, a classification unit classifying the subject pixel based on the class taps, a coefficient output unit outputting a coefficient corresponding to a class of the subject pixel from among coefficients, determined beforehand by learning, corresponding to the plurality of classes, and a calculator determining the subject pixel by performing prediction calculations by using the coefficient corresponding to the class of the subject pixel and the prediction taps.

Abstract: This invention discloses methods and apparatus for ultrasonic color imaging that characterizes ultra-fine structures and distributional physical conditions within the target under inspection. The disclosed complementarily incorporates information arising from a plurality of repeated sound trips forced by repeated reflections of exterior and interior interfaces of target, and expresses the information into an image segment representing the main path that ultrasonic signals traveled within the target. The image produced is substantially more discriminative, descriptive, and position-sensitive to both acoustic interfaces and distributional acoustic characteristics of the target. The invention is especially useful for thin sheet targets most vulnerable to both non-continuous and continuous interior defects.

Abstract: The disclosed is to provide an image signal processor and its method in which the enhancement of resolution and the effective reduction of noise are compatible. The image signal processor is provided with an edge character detector that detects a character of an edge of an object, an image signal frequency zone limiter that executes a noise removing process, an image signal frequency zone enhancer that executes a resolution enhancing process and a gain controller that changes each intensity of the image signal frequency zone limiter and the image signal frequency zone enhancer according to an output result of the edge character detector.

Abstract: A system and method for enhancing the contrast within an image. An enhanced image can be generated in a real-time or substantially real-time manner from an initial image. A low-contrast image is enhanced, including the visibility of text within the low-contrast image. A logarithm transform method is used to enhance the overall contrast and visibility within the image. A power transform method is used to enhance the visibility and contrast of written text, symbols and other forms of signs within the enhanced image.

Abstract: Methods and apparatus are provided for visual background subtraction using one or more preprocessing modules. One or more effects are detected in a received image signal and one or more blocks are selectively enabled to preprocess the image signal to compensate for the detected one or more effects. Visual analysis is then performed on the preprocessed signal using background subtraction. A spatially-variant temporal smoothing of the image signal is also disclosed. The spatially-variant temporal smoothing can be achieved by the mixing of a new intensity value with a previous intensity time-average as determined by a weighting matrix. The mixing can be influenced by a dynamic bias term that is a real-time estimate of a variance at the pixel, such as a degree of change, and the weighting can be determined by a relative stability of an observed value compared to a stability of the time-average.

Abstract: An image-processing method includes: subjecting original image data indicative of an original image to a selective noise reduction by reducing noise from pixel data that meets a prescribed condition among all the pixel data in the original image data; and performing a first calibration to calibrate contrast of a backlight region of the selectively-noise-reduced original image data to generate an output image by determining a first pixel component and a second pixel component for each pixel data in the selectively-noise-reduced image data, by increasing contrast of the second pixel component, and by recreating pixel data based on the first pixel component and the calibrated second pixel component.

Abstract: A method for identifying and determining the frequency of scattered radiation fringe patterns for applications such as particle or droplet sizing and laser Doppler velocimetery. The method utilizes a series of windowed Fourier transforms performed on an intensity profile of the scattered radiation to locate a segment where a fringe pattern is located on the intensity profile. A standard Fourier transform is then performed on the segment to determine a dominant frequency of the fringe pattern, from which a physical quantity such as diameter or velocity of the particle or droplet may be derived. The method may be utilized with a line or an array sensor to measure the fringe patterns in a spatial domain, or with a point detector to measure the fringe patterns in the time domain.

Abstract: A stereo imaging based vision system and method provides enhanced portal security through stereoscopy. In particular, a system detects non-people objects within the chamber of a revolving door by acquiring two-dimensional (2D) images from different vantage points, and computing a filtered set of three-dimensional (3D) features of the door compartment by using both the acquired 2D images and model 2D images. Applying image processing techniques to the filtered 3D feature set, non-people objects can be detected.

Abstract: A method of image processing using kernel regression is provided. An image gradient is estimated from original data that is analyzed for local structures by computing a scaling parameter, a rotation parameter and an elongation parameter using singular value decomposition on local gradients of the estimated gradients locally to provide steering matrices. A steering kernel regression having steering matrices is applied to the original data to provide a reconstructed image and new image gradients. The new gradients are analyzed using singular value decomposition to provide new steering matrices. The steering kernel regression with the new steering matrices is applied to the noisy data to provide a new reconstructed image and further new gradients. The last two steps are repeated up to ten iterations to denoise the original noisy data and improve the local image structure.

Abstract: In an object recognition apparatus, a first division unit and a second division unit each partitions an image into a plurality of regions. A first calculation and a second calculation unit each derives, for each of the plurality of regions partitioned. A first comparison unit and a second comparison unit each compares the derived characteristic quantities in between at least two images, for each of the plurality of regions. A recognition unit recognizes a region where the object is located, based on the comparison result.

Abstract: A method for aligning a pair of images includes providing a pair of images, identifying salient feature regions in both a first image and a second image, wherein each region is associated with a spatial scale, representing feature regions by a center point of each region, registering the feature points of one image with the feature points of the other image based on local intensities, ordering said feature pairs by a similarity measure, and optimizing a joint correspondence set of feature pairs by refining the center points to sub-pixel accuracy.

Abstract: A method and apparatus to receive an image and to create a smaller representation of the image from a wavelet representation of the image is described. The size of the smaller representation of the image is selected to compensate for the content of the image and physical properties of the display device to display the smaller representation of the image.

Abstract: An interactive photo system that provides an interesting and entertaining way to link photos from the digital and analog domains. The system enables easy conversion of a paper-based photo into digital form. The conversion process generally involves a user holding up a paper-based photo image in front of a camera of the system, such that the captured image is rendered in a specific location on a display screen. The system adjusts the captured image for orientation and illumination irregularities in creating a digital version of the paper-based photo. The system further includes a database of digital images and a content retrieval engine to which content from the just-converted digital image can be used as a search query to search the database for other similar digital photos.

Abstract: An image storage device according to an example of the invention comprising a first storage unit which stores first image, a second storage unit which stores second image to be determined whether the second image includes detection target data which is the similar to data included in the first image, parallel comparison units which perform a process of comparing pixel values between a frame of the first image and a frame of the second image, and determining, based on a comparison result between the frame of the first image and the frame of the second image, whether the detection target data is included in the second image in parallel, and a third storage unit which stores, when it is determined that the detection target data is included in the second image, the detection target data.