Abstract: An aspect of the present invention provides a moving obstacle detecting device installed in or on a moving body for detecting moving obstacles in the vicinity of the moving body, the moving obstacle detecting device that includes an image capturing unit configured and arranged to capture an image of the surroundings of the moving body, an optical flow calculating unit configured to find an optical flow corresponding to characteristic points in the image captured with the camera, a focus of expansion calculating unit configured to calculate focus or foci of expansion of the motion vectors in the optical flow found by the optical flow calculating unit, a histogram creating unit configured to create a histogram of the distribution of the focus or foci of expansion calculated by the focus of expansion calculating unit, a peak test unit configured to detect at least one moving obstacle to test peaks in the histogram, the test determining whether or not the peaks correspond to the same object if two or more peaks e

Abstract: An image input unit inputs an image including a user's face area to specify the user's gaze or face direction. A pattern generation unit generates an input pattern that can be compared with each dictionary pattern from the image. A dictionary unit previously stores a plurality of dictionary patterns each of which differently corresponds to the user's gaze or face direction. A pattern comparison unit respectively calculates a similarity degree between the input pattern and each of the dictionary patterns by comparing the input pattern with each of the dictionary patterns. A gaze detection unit detects the user's gaze or face direction based on the similarity degree.

Abstract: An object of the invention is to provide a target travel direction detecting method capable of detecting the travel direction of efficiently detecting the travel direction of a target by using a small amount of information. The target travel direction detecting method according to the invention generates a lightness image from each of the two color images picked up at predetermined time intervals and causes a template of n by n 105 to scan the lightness image at time t to detect a position where the target is present. Next, the values of pixels around the pixel of the position where the target is detected are stored into the template 105. The template 105 storing the pixel values is used to scan the lightness image at time t+1 and detect a pixel that matches each pixel value of the template 105. Then, the target is plotted for each travel direction at a pixel position matching the template 105 on the lightness image at time t+1 to create a travel-direction-extracted image.

Abstract: A reference posture and a reference position of the imaginary sphere and an arbitrary position are set. The relationship between three-dimensional coordinates and two-dimensional coordinates is derived by using a photographing means. Positions of marks given to the imaginary sphere formed at the coordinates in the three-dimensional space are converted into positions on a two-dimensional image. An operation of displacing a posture of the imaginary sphere relative to the reference posture and the reference position is performed in such a way that the coordinate values of the two-dimensional imaginary marks and the coordinate values of the marks present on the two-dimensional images of the sphere are coincident with each other. The rotational and flight characteristics of the sphere are computed, according to the three-dimensional posture and position at one time and the three-dimensional posture and position thereof at another time.

Abstract: Biometric identification and/or authentication method for identifying and/or authenticating the user (1) of a terminal (2), wherein said user is identified and/or authenticated by means of personal involuntary eye movement patters being determined with an eye-movement detection module (20). The user is identified and/or authenticated for example by means of characteristics of the personal trajectory (51) of the eyes on an image being determined.

Abstract: A surveillance camera and illuminator are covertly mounted in a partially rotatable bay window inner frame mounted within an electrical switch box. The bay window inner frame can thus be pivoted back and forth to orient the camera toward an entrance to a room in which the switch box is mounted. When the bay window inner frame is rotated to the desired position, a bay window switch box decor cover plate is mounted over the bay window inner frame, engaging it and locking it in the selected position by attachment of the cover plate to the switch box.

Abstract: In a method of improving fingerprint images, wherein recorded images are processed on image areas of the fingerprint images by using Gabor filters, the Gabor filters are adapted to characteristic quantities of the image in the relevant image area to be processed.

Abstract: An inter-image operation apparatus has temporal subtraction means for calculating a temporal subtraction image from two or more sets of base image data, and history data recording means for recording history data concerning the calculated temporal subtraction image in a database server and concurrently attaching the same history data to the data on the temporal subtraction image. The history data may include information identifying the two or more base images, the date of calculating the temporal subtraction image, the ID number of the patient, the name of the photographed portion of the patient, the photographing angle of the patient, etc. A desired temporal subtraction image may be immediately fetched from the database server by referring to the history data attached thereto if that temporal subtraction image had already been obtained, instead of recalculating that temporal subtraction image. Thus, the diagnostic process of comparing two or more images of an identical object can be carried out efficiently.

Abstract: A method for providing image correlation between the images displayed on a PACS system, the method including the steps of: displaying a first image from a first stack of images and a second image from a second stack of images, each stack of images belonging to a different series of image captures for a subject from a particular orientation; determining a spatial location of each image in the first and second stacks of images; receiving a selection signal for an image in the first stack of images; determining a corresponding image in the second stack of images based on the selection signal and the spatial location of the corresponding image; and, displaying the corresponding image. An apparatus for displaying correlated images based on anatomic triangulation.

Abstract: A computer controlled method for detecting and diagnosing a rare cell type in a tissue sample is provided, said method comprising treating the tissue sample such that it generates a first signal indicative of the presence at a location of a rare cell, detecting the first signal, treating the location at which the first signal is detected to generate a second signal indicative of a diagnostically useful cellular characteristic and detecting the second signal. The first signal can be morphological or a color present in a sought cell either before or after staining. The second signal can be generated by in situ PCR or PCR in situ hybridization. In one preferred embodiment, the rare cell type is fetal cell in a maternal blood tissue sample, said sample consisting of a smear of unenriched maternal blood. In another embodiment, the method is used to diagnose or genotype cancer cells in a blood or tissue biopsy sample.

Abstract: Structures are delineated in medical or other images. First, various tissue types present in the image are statistically described using a maximum likelihood classifier. Second, the tissue of interest is described using an exemplar, which is derived either from an anatomical atlas or from user input. Third, the structure of interest is morphologically described. The process can be iterated until a desired level of accuracy is achieved.

Abstract: A CAD system pre-analyzes a medical image to automatically determine whether contrast agent was introduced into the patient before forming that image. In one embodiment, contrast agent detection is performed by comparing the voxel values in known blood-containing areas to voxel values of known water-containing areas, and if there is a substantial difference above a preferred threshold, then it is determined that this patient was indeed injected with the contrast agent. Upon finding that the patient was indeed injected with contrast agent, the CAD algorithm then automatically performs both a detection algorithm that uses the contrast agent and one that does not and notifies the clinician that both sets of results are available to look at. However, if it is determined that the patient was not injected with contrast agent, the CAD algorithm does not perform the algorithm that uses the contrast agent does not give the user the option to view any results from such algorithm.

Abstract: A method for detecting, quantifying, staging, reporting, and/or tracking of a disease includes providing analysis software configured to detect, quantify, stage, report, and/or track a disease utilizing images of a patient. The analysis software is executable on a personal computer of a patient. Patients are then imaged utilizing a medical imaging apparatus and medical images of the patient produced by the imaging apparatus are downloaded to the personal computer of the patient. The imaging and downloading are repeated a plurality of times at intervals selected to provide the analysis software with sufficient images to detect, quantify, stage, report, and/or track the disease in the patient.

Abstract: A phase contrast image is generated on the basis of a plurality of radiation images of an object taken in imaging positions which are different from each other in the distance from the object. An enlargement/reduction processing is carried out on the radiation images as taken in the imaging positions according to the distances between the imaging positions so that the radiation images become substantially the same in their sizes, and a phase contrast image is generated on the basis of the radiation images thus processed.

Abstract: The present invention achieves rapid identification of pathogens, bacteria, cancer cells and other abnormal human and animal cells. In one embodiment, the system of the present invention comprises a first subsystem that obtains and processes images of specimens of pathogens, bacteria, and other abnormal cells, and a second subsystem that accepts the images, isolates the particular features of the image using advanced image segmentation, and then rapidly and accurately identifies the pathogens, bacteria and other abnormal cells by using a pattern recognition process wherein the segmented or isolated features of the original image are compared to known reference images.

Abstract: An optical object detector including a feed path, a reference pattern facing the feed path, and an optical sensor configured to view the reference pattern through the feed path, absence of the reference pattern from the view of the optical sensor indicating presence of an object in the feed path.

Abstract: In an image defect inspection method and apparatus which detects a gray level difference between corresponding portions of two images, automatically sets a threshold value based on the distribution thereof, compares the gray level difference with the threshold value, and judges one or the other of the portions to be defective if the gray level difference is larger than the threshold value, provisions are made to correct the threshold value when the distribution of the gray level difference is different from the usual distribution, thereby achieving high detection sensitivity while suppressing the occurrence of false defects.

Abstract: A neural network is trained and used to reduce artifacts in spatial domain representations of images that were compressed by a transform method and then decompressed. For example, the neural network can be trained and used to reduce artifacts such as blocking and ringing artifacts in JPEG images.

Abstract: A detection and classification method of a shape in a medical image is provided. It is based on generating a plurality of 2-D sections through a 3-D volume in the medical image. In general, there are two steps. The first step is feature estimation to generate shape signatures for candidate volumes containing candidate shapes. The feature estimation method computes descriptors of objects or of their images. The second general step involves classification of these shape signatures for diagnosis. A classifier contains, builds and/or trains a database of descriptors for previously seen shapes, and then maps descriptors of novel images to categories corresponding to previously seen shapes or classes of shapes.

Abstract: An image processing device according to the present invention includes a similarity factor calculating portion locally comparing pixels that compose an image so as to calculate a similarity factor of the pixels in a predetermined direction, and includes an analyzing portion analyzing the structure of the image in accordance with the similarity factor calculated by the similarity factor calculating portion. In particular, the similarity factor calculating portion selects and changes pixels to be compared in accordance with color shift of the image to calculate the similarity factor. As a result, a similarity factor can be calculated with suppressing influence of color shift.

Abstract: A method and system for classifying images between natural pictures and synthetic graphics is provided. In embodiments of the invention, a picture/graphic classification method and system implements one-dimensional color discreteness features, two-dimensional color discreteness features, or three-dimensional color discreteness features for image classification of natural pictures and synthetic graphics. In another embodiment of the invention, a picture/graphic combination classification method and system implements one-dimensional color discreteness features, two-dimensional color discreteness features, and/or three-dimensional color discreteness features is used to classify images between natural picture and synthetic graphic classes.

Abstract: A method is disclosed of analyzing an image to identify “must exclude” regions and of determining the best fit of an image outline so as to generate an image boundary that excludes the majority of “must exclude” regions. The method is applicable to the automatic generation of greeting cards and photo-montages.

Abstract: A method for optically recognizing alphanumeric characters which involves storing polygonal outer contours and inner contours of reference characters and their relative geometrical positions as normalized reference curves transformed into a tangent space. Outer and inner contours of a character to be recognized are optically scanned and the scanned data is converted into binary data. The outer and inner contours of the character to be recognized are approximated as a polygon, and straight line segments of the polygon are normalized and transformed into a tangent space to generate character curves. The character to be recognized is compared to the reference characters in a rotation-invariant manner by shifting at least its character curve for the outer curve and the reference curves of the outer contours relative to each other in the tangent space until a surface area enclosed by the character curve and a given reference curve is at a minimum.

Abstract: In a method for capturing a complete data set of forms of identical layout provided with graphic characters in separate data fields, an image of the form is produced and image data are saved. Aided by a character recognition program, the characters of the data fields are identified as much as possible according to a predetermined degree of certainty. Image data of unidentified data fields are correlated with a reference parameter to generate a transfer data set that is sent to an external evaluation station where the characters of the unidentified data field are identified based on image data of the transfer data set. The identified characters are combined to a character data set and compiled with the reference parameter to a processed transfer data set and returned. Based on the reference parameter, the character data set is correlated with the other data fields to generate a complete data set.

Abstract: An apparatus for capturing a document with a plurality of compression models comprises an image-processing system and an image-capturing system. The image-processing system is for accepting a request of setting the compression models for the document and processing the data of the document according to the compression models, wherein the request is from an exterior electric device connected to the apparatus. The image-capturing system is for capturing the data of the document, and is connected to the image-processing system and comprises multitudes of optic devices. The method for capturing the document with these compression models comprises accepting a request of setting the compression models for the document; capturing the data of the document; and processing the data of the document with these compression models.

Abstract: A video processing system 100 is provided with video encoding device 10 and video decoding device 20. The video encoding device 10 implements backward interframe prediction from a temporally subsequent frame and outputs information indicating that an option to eliminate use of a decoded image of the temporally subsequent frame was chosen. The video decoding device 20 eliminates use of the decoded image of the frame on the basis of this information, in conjunction with input of the information for eliminating use of the decoded image of the temporally subsequent frame.

Abstract: A method and Electro-Optical (EO) system for processing imagery comprising selecting a first frame of data as a template frame; capturing a second frame of data using the EO system, for a plurality of pixels of the second frame, correlating the plurality of pixels of the second frame with pixels of the template frame to generate a plurality of shift vectors, one for each pixel of the plurality of pixels of the second frame, registering the second frame with the template frame by interpolating the second frame using the plurality of shift vectors and re-sampling at least a portion of the second frame to produce a registered frame, re-sampling the template frame, and combining the re-sampled template frame and the registered frame to generate an averaged frame.

Abstract: An aspect of the present invention includes a method of encoding coefficients of a subband decomposition, to prepare the coefficients for compression. Another aspect of the present invention includes an inverse method of decoding coefficients of a subband decomposition, after they have been decompressed. Particular features and other aspects of the present invention are indicated in the specification, including the drawings and claims.

Abstract: The invention concerns a method for encoding at least a source image using a hierarchical mesh defining at least two nested spaces corresponding each to a decomposition level of said meshing, which consists, at least at a decomposition level n (except the first decomposition level), in delivering only image coefficients expressed in a base of functions defined in a space orthogonal to the space corresponding to the preceding decomposition level n?1, said functions being selected so that said image coefficients enable to optimize for said decomposition level n the data already transmitted for the preceding decomposition level n?1, so as to produce a reconstructed image, representing said source image, with optimized restoration quality for said decomposition level n.

Abstract: The invention is directed to a method and apparatus for reducing the bandwidth required to transmit image data to a display device. It is contemplated that the image data represents a plurality of pixels defining an image for display on the display device, and that the image data is transmitted in a data stream with the data being in a predetermined order of pixel components. The method comprises the steps of receiving the image data from the data stream, and transforming the image data as it is received by selectively storing some of the image data in a memory for access by the display device and discarding other of the image data.

Abstract: A system and method for producing formatted information related to at least one image-capture appliance and/or at least one image-restitution appliance. The system and method provide for production of formatted information related to defects of the appliance, especially blurring and noise, from a measured field. This measured field is obtained from a universal set by performing the operations: capturing or restituting the universal set by the appliance to produce an image; choosing an image zone within the image; calculating a reference image from the image; selecting, within the reference image, a reference zone homologous to the image zone; choosing a base for representing all or part of the image zones and reference zones; selecting variable characteristics of the appliance; and in this way the measured field is obtained. The system and method are applicable to photographic or video image processing, in optical devices, industrial controls, robotics, metrology, etc.

Abstract: A method and apparatus for object-of-interest image de-blurring includes functions that determine (305, 310) an object region of an object-of-interest in an image that may be generated by image capture device (805), that determine (315, 825) a motion vector of the object-of-interest, that determine (325, 830) a scaling factor, and that generate (330) a de-blurred object region by processing the object region using a deconvolution filter (835) formulated from the motion vector and the scaling factor.

Abstract: An image processing apparatus includes a filtering processor that includes a filter and switches between a pre-filtering path and a post-filtering path; a data compressor that compresses image data output from the filtering processor; an image data storage that stores the image data compressed; and a data expandor that expands the image data stored. The pre-filtering path includes a first path sending a raw image data to the data compressor via the filter and a second path passing the image data expanded. The post-filtering path includes a third path passing the raw image data to the data compressor and a fourth path sending the image data expanded to the filter.

Abstract: A method classifies pixels in an image by first partitioning the image into blocks. A variance of an intensity is determined for each pixel, and for each block the pixel with the maximum variance is identified. Then, the blocks are classified into classes according to the maximum variance.

Abstract: Demosaicing of graphical content is provided. In an illustrative implementation a demosaicing engine executing one or more demosaicing algorithms is employed to operate on graphical content to provide better quality and higher resolution images. In operation, the demosaicing engine operates in two modes, a training/learning mode, and a run time mode. During training, training-images are analyzed to generate a codebook of mosaic filter table entries, such that each table entry has an associated list of similar training pixel blocks and their associated filters. During run time, a run-time image is broken into pixel blocks. Each pixel block is then compared with the entries of the codebook to find the closest match filter. The list associated with the entry is then processed using a least-squares algorithm to locate the optimal mosaic filter. As a result, higher resolution is achieved without requiring more pixels.

Abstract: Disclosed herein are methods and apparatus, including computer program products, implementing and using techniques for reducing mosquito noise in an image frame of a video signal. A window of the image frame is provided. The window includes a plurality of pixel values. One of the pixel values is selected from the pixel values in the window to define a modified window of pixel value differences. A local mean (DC) value is computed based on the pixel value differences. A mosquito noise reduction value is also computed based on the pixel value differences. An artifact attenuation factor is computed as a function of a dynamic range measure of the pixel value differences. A weighted sum of the mosquito noise reduction value and the local mean value, applying the artifact attenuation factor, and the selected pixel value, is computed to generate a modified pixel value.

Abstract: Biometry for transforming, receiving, processing and analyzing electronic images of living biological objects is provided. The Biometry detects the objects, which produce insignificant periodical oscillations thereby making it possible to receive and analyze the images of the objects on a background of stationary and moving objects. Successive exposures of the object's image are received and picture-to-picture differences are evaluated, and a sum of differences of at least two successive exposures during the processing thereof is accumulated. The biometry comprises a photosensitive multielement transducer (1) and means for processing the images of the picture-to-picture difference which contain an exposure memory (2,3), a unit for evaluating the picture-to-picture difference (4), a picture-to-picture difference summator (5) and control registers (6). The device accumulates the sum of the exposure differences detected by the photosensitive multielement transducer.

Abstract: One or more embodiments of the present invention relates to a method and system for simultaneously generating spatially compounded and non-compounded images. In one embodiment, the present invention relates to a method for generating images using an ultrasound apparatus. This method comprises storing at least one frame and generating at least one image output from the at least one frame. The method further comprises displaying at least the generated image output.

Abstract: Ink is stored in a data structure that includes a mapping of the ink's coordinates from one space to another. The responsibility of transforming the ink into desired coordinates occurs after the capture of the ink and at least based on the mapping in the data structure.

Abstract: Ink is stored in a data structure that includes a mapping of the ink's coordinates from one space to another. The responsibility of transforming the ink into desired coordinates occurs after the capture of the ink and at least based on the mapping in the data structure.

Abstract: The present invention relates to an image processing device including pixel interpolation, such as a pixel interpolation device, a line interpolation device, an image size conversion device and an image interpolation device. The pixel interpolation device according to the present invention includes an interpolation pattern table, which outputs an interpolation direction data designating interpolation directions including directions joining intermediate points of pixels. The interpolation pixel data is calculated based on the pixel data located in the interpolation direction designated by the interpolation direction data. The image interpolation device, the pixel number conversion device and the line interpolation device according to the present invention includes the pixel interpolation device described above.

Abstract: The field of the invention is that of the semiconductor optical devices used in particular for fibre-optic telecommunications. To function efficiently, a certain number of semiconductor devices require the use of light polarized in a given polarization state. When knowledge of the polarization the state is lost, the optical element according to the invention makes it possible to polarize the light again in a known polarization state. By using two of these elements in combination with a coupler, it is possible to produce a device which fulfils the same function as a polarization splitter. This optical assembly delivers two output signals whose polarization states are the projections of the initial polarization onto two orthogonal axes. The main advantage of these devices is that they are produced using polarization rotators based on photonic crystals, and they can consequently be integrated easily into semiconductor devices, which the use of discrete polarizers does not allow.

Abstract: An optical reader system is described herein which has a single mode (SM) optical fiber launch/receive system that uses one or more SM optical fibers to interrogate a biosensor and does not use multimode (MM) optical fibers to interrogate the biosensor. The use of the SM optical fiber launch/receive system effectively reduces angular sensitivity, reduces unwanted system reflections, improves overall angular tolerance, and improves resonant peak reflectivity and resonant peak width. Two specific embodiments of the SM optical fiber launch/receive system are described herein which include: (1) a dual fiber collimator launch/receive system; and (2) a single fiber launch/receive system that interrogates the biosensor at a normal incidence.

Abstract: Effects of diffraction of a spectral beam from an edge of the micromirrors are reduced in order to optimize the passband in a wavelength selective switch. The effects of diffraction on the pass band may be reduced by using rotation of the micromirror about both the attenuation axis and the switching axis to achieve the desired level of attenuation. Peak coupling can be attained by dithering the micromirror about an axis tangent to a contour of constant attenuation using simultaneous rotation about the switching and attenuation axes.

Abstract: A cylindrical lens array cannot be manufactured so that each cylindrical lens has the same radius of curvature and the same accuracy in the surface. Therefore, when the laser annealing is performed using the cylindrical lens array, the beam spots divided by the cylindrical lens array cannot be superposed completely in the same surface. As a result, there is a region where the energy is attenuated in the edge portion of the rectangular beam to be formed, and therefore the intensity distribution of the laser beam becomes inhomogeneous. In the present invention, the cylindrical lens array is used in combination with the optical waveguide. After dividing the laser beam in a predetermined direction by the cylindrical lens array, the divided beams are combined, and then the laser beam is incident into the optical waveguide that acts upon the same direction as the predetermined direction. This can correct the variation in the intensity of the laser beam due to the processing inaccuracy of the cylindrical lens array.

Abstract: A collimator lens having a low reflectivity and significantly improved durability against high-power light is provided so that optical parts using the collimator lens can be reduced in size and cost lowered. A plurality of optical fibers can be connected to the collimator lens to produce a variety of fiber collimators and optical parts based on the fiber collimator. The collimator lens is made essentially of quarts glass whose refractive index is graded radially so as to increase towards the optical axis and decrease gradually towards the outer periphery. The collimator lens and optical fibers are connected directly by fusion.

Abstract: An optical fiber coupling part capable of reducing coupling loss while maintaining a large operating distance, and having a good module assembling property. AT least one GRIN lens having numerical aperture NA that is larger than numerical aperture NAs of a light-emitting source (such as a semiconductor laser) is fusion-spliced with one end of the optical fiber. All lights emitted from the light-emitting source can enter the GRIN lens, and the loss of the light can thereby be reduced. In addition, a second GRIN lens having numerical aperture NA2 is fusion-spliced with one end of the optical fiber having numerical aperture NAf, and further a first GRIN lens having numerical aperture NA1, which is larger than numerical aperture NA2, is fusion-spliced with the other end of the second GRIN lens. Thereby, the light emitted from the light-emitting source can efficiently enter the optical fiber, and loss of the light can thereby be reduced. In this case, the formula expressed by NAf?NA2<NAs?NA1 is desirable.

Abstract: A high index difference coupler includes a high index difference waveguide having one or more modes. A plurality of gratings is formed on the high index difference waveguide. The effective index difference between low index regions and high index regions of the waveguide is greater than 0.3.

Abstract: A device for crossing optical beams. At least a first input optical waveguide is directed along a first axis, a second optical waveguide is directed along a second axis inclined with respect to the first axis, and a photonic crystal having a regular periodicity in an optical crossing region at the intersection of said first and second axis has, opposite the first and second input waveguides, a first and a second output waveguide to provide for exit beams.

Abstract: Integrated fiber tap monitors with variable optical attenuators where light from an input fiber is reflected to the output fiber and a tap is generated from measuring transmission at the reflection.