Abstract: The invention is a method for detecting events in an imaged scene by analyzing the occlusion of linear features in the background image. Linear features, curved or straight, in specific scene locations are either manually specified or automatically learned from an image or image sequence of the background scene. For each linear feature, an occlusion model determines whether the line or part of it is occluded. The locations of the lines of interest in the scene, together with their occlusion characterizations, collectively form a description of the scene for a particular image. An event, defined as a series of descriptions of the scene over an image sequence, can then be initially defined and subsequently detected automatically by the system. An example application of this is counting cars or people passing in front of a video camera.

Abstract: According to an asymmetric watermarking technology, which is particularly resistant to projection attack, an original image is analyzed to obtain its watermarking space. The watermarking space is then divided to obtain two orthogonal subspaces g and h. An embedding key G, which is a matrix and which columns form bases of subspace g, is selected and a matrix H, HTG=0 is calculated. Columns of matrix H form bases of subspace h. A detecting key D, which equals the sum D= GT=BHT, wherein B is a matrix, is calculated, and a watermark w is obtained and embedded into the original image ?0 to obtain a watermarked image ?w=?0+Gw, wherein D?0=m0 is not a 0 vector.

Abstract: The present invention relates to a method for assessing the presence or absence of lesion(s) in an image and a system therefor, wherein said image may be any image potentially comprising lesions, in particular an image from medical image diagnostics, and more particularly an ocular fundus image. The lesions are identified from starting points being candidate lesion areas and validated with respect to their visibility as compared to the local surroundings.

Abstract: A method for processing an image of a microfluidic device. The method includes receiving a first image of a microfluidic device. The first image corresponds to a first state. Additionally, the method includes receiving a second image of the microfluidic device. The second image corresponds to a second state. Moreover, the method includes transforming the first image and the second image into a third coordinate space. Also, the method includes obtaining a third image based on at least information associated with the transformed first image and the transformed second image, and processing the third image to obtain information associated with the first state and the second state.

Abstract: A method and system for collecting traffic data are disclosed. From a satellite or aircraft located at an altitude above a geographic area, a thermal image of the geographic area is obtained. Using a geographic database, the thermal image is matched to the positions of roads in the geographic area. The thermal image is used to determine traffic parameters, such as densities and speed, at specific locations along the roads.

Abstract: A device and method for reading and authenticating currency serial numbers to prevent counterfeiting and other money-related crime is provided. The device may be retrofitted to be installed in an existing money counting machine, and may read serial numbers of currency that passes through the counting machine. The serial numbers can be captured from stationary currency prior to entering the counting machine, and the serial numbers may be authenticated by comparing the number against various databases.

Abstract: A method and apparatus for calibrating an apparatus that acquires a sequence of radiographic images and correcting images of an object under observation. For each image of a sequence acquired by the apparatus and for a given frequency of acquisition of the sequence, the apparatus is calibrated by determining the value of the variation of a mean of gray levels in at least one zone of interest of the current image of at least one calibration device, the variation being determined relative to the mean gray level of the first image of the sequence in each zone of interest. The determination of the variation is reiterated for a series of images sequences acquired using calibration devices resulting in first images of mean gray levels different from one sequence to another.

Abstract: In an apparatus for processing banknotes or other sheet-like objects, the banknotes are conveyed along a transport path, thereby passing sensor means. A transmission and a reflection image of the banknote are captured by illuminating the banknotes and measuring the transmittive and reflective light intensity of light with a high resolution. Based on these images, a validation of the banknotes is carried out. After the validation, a double feed detection is performed by evaluating the transmission and the reflection intensities for a predefined set of test spots with a two-dimensional evaluation. If it is decided on double feed, the banknotes are rejected. Otherwise the banknotes are accepted and further processed. Because of the additional dimension in the evaluation compared with known double detection methods, the invention enables a more robust double note detection also in cases with different degrees of soiling of the banknotes.

Abstract: The present invention relates to a method and a system for synthesizing an intensity pattern based on generalized phase contrast imaging. The phase filter contains a plurality of phase shifting regions that is matched to the layout of a light source array, each of the regions being positioned at the zero-order diffraction region of a respective element of the array. Further, the shape of each phase shifting region may match the shape of the zero-order diffraction region of the respective element. Thus, the energy of the electromagnetic fields of the system may be distributed over a large area compared to the area of a zero-order diffraction region of a single plane electromagnetic field of a known phase contrast imaging system.

Abstract: An action analysis method comprising: an image recording step for continuously photographing a target; a movement locus measuring step for obtaining a movement locus of the target; an action analysis step for analyzing an action of the target from the movement locus; and an output step. In the action analysis step, a speed v and a movement direction displacement angle ? for each unit time is obtained from the movement locus; an average value X of v·cos ? for each unit time within a predetermined time is obtained; and a predetermined extraction target is distinguished by a statistic measure using the average value as an operation element index.

Abstract: An apparatus and a method for compressing a picture by wavelet transform and entropy coding, in which the processing load of arithmetic coding may be relieved and rate control may be carried out effectively, while still maintaining the picture quality. An encoding object prediction unit predicts bitplanes to be encoded and extracts only those bitplanes which are to be encoded. An EBCOT unit generates an encoding pass from one bitplane to be encoded to another to execute arithmetic coding. In controlling the code volume to a target code volume, a rate controller sums the arithmetic codes in a sequence from the arithmetic code with the highest bit position in the totality of the code blocks of an input picture to the arithmetic code of the lowermost bit, from one bitplane to another or from one encoding pass to another, and halts the summation when the predetermined target code volume is exceeded.

Abstract: A fundamental frequency and at least one harmonic frequency of gridlines in a radiographic image are measured to form a filter for extracting the measured fundamental frequency component and harmonic frequency component. Filtering is performed on the radiographic image using the formed filter. Based on the obtained image data, a gridline model which takes the measured fundamental frequency and harmonic frequency into consideration is fitted to generate a grid image. The gridlines are removed from the radiographic image by the generated grid image.

Abstract: A digital watermark embedding apparatus comprises below units. A first generation unit generates a plurality of symbol sequences each of which includes a plurality of symbols including ranks, each of the ranks being uniquely numbered among each of the symbol sequences, each of the symbol sequences uniquely corresponding to each of a plurality of identification information items to be embedded as digital watermark information into each of copies of digital contents. A second generation generates a plurality of to-be-embedded codes corresponding to each of the symbols in each of the symbol sequences. An embedding unit embeds the to-be-embedded codes in each of the copies.

Abstract: A method of identifying errors in information describing terrain features is disclosed. According to the method, a first set of data is obtained which describes an area of terrain. The first set of data has a first type of error. A second set of data is obtained which describes the area of terrain. The second set of data has a second type of error. The second type of error is characteristically different from the first type of error. A subset of the first set of data corresponding to a first portion of the area of terrain is compared with a subset of the second set of data corresponding to the first portion to identify information not present in one of the first and second sets of data.

Abstract: A real-time contents editing system edits a large number of images and/or voices which are present in a dispersed fashion on the Internet, and distributes the edited images and/or voices to a plurality of users. The system includes a plurality of video cameras, a plurality of distribution modules each adapted to code an input image taken by a corresponding video camera by use of a coding standard which enables coding while selecting one of a plurality of coding algorithms and to distribute the coded input images, a plurality of receiving modules each adapted to receive and display the images, and at least one editing module. Each distribution module changes, in accordance with the performance level of a machine to be used, the kind and use frequency of a video object plane (VOP) to be used, to thereby select a coding algorithm which enables highly efficient compression.

Abstract: One embodiment disclosed relates to a method for encoding and decoding a video sequence in which a keyframe is used to bi-directionally predict frames in the sequence. The keyframe is coded independently of other frames in the sequence. Both a prior frame occurring before the keyframe and a subsequent frame occurring after the keyframe are predicted using data from the keyframe. Another embodiment disclosed relates to a method for allocating bits to a keyframe during video encoding. Effects of a plurality of keyframe bit allocations on quality of a predicted frame are measured. Said effects are used to determine a near optimal keyframe bit allocation.

Abstract: A method of detecting lesions and polyps in a digital image, wherein said image comprises a plurality of 3D volume points, is provided. The method includes identifying a surface in the image, and for each point in the image, calculating a first curvature measure, forming a set of rings of points about each point, each ring being of equal geodesic distance from a center point for the ring, calculating, for each ring, a standard deviation of the first curvature measure, and selecting those rings with a minimum standard deviation for the first curvature measure. A first curvature slope is calculated for the selected rings, those points where the curvature slope departs from the pattern expected for a polyp or lesion are deleted from the surface.

Abstract: A personal authentication apparatus includes a facial region extraction which extracts an image of a facial region of a person obtained from image sensing input unit, a guide unit which guides motion of the person of interest, a feature amount extraction unit which extracts the feature amount of a face from the image of the facial region extracted by the facial region extraction unit while the motion is guided by the guide unit, a dictionary registration unit which registers the feature amount extracted by the feature amount extraction unit as a feature amount of the person of interest, and an authentication unit which authenticates the person of interest in accordance with the similarity between the feature amount extracted by the feature amount extraction unit, and a feature amount which is registered by the dictionary registration unit.

Abstract: An image input device inputs image sequence data. A memory stores each image of the image sequence data. A tracking point indication unit initially indicates a plurality of tracking points on a first image stored in the memory. A pattern matching unit compares a pattern of each tracking point of the first image with a second image stored in the memory, and detects a plurality of destination candidate points from the second image, each corresponding to each tracking point. A position inconsistency correction unit decides whether a positional relation of the plurality of destination candidate points is consistent with a positional relation of the plurality of tracking points, and corrects at least one of the plurality of destination candidate points on the second image if the positional relation of the plurality of destination candidate points is inconsistent with the positional relation of the plurality of tracking points.

Abstract: A method and system for mutually immersive telepresencing are provided. A user is viewed at a user's location to provide a user's image. The size of the user's head is determined in the user's image. A surrogate having a surrogate's face display about the size of the user's head is provided. The user's image is processed based on the size of the surrogate's face display to provide an about life-size image of the user's head. The about life-size image is displayed on the surrogate's face display.