Abstract: Embodiments of a computer-implemented method for training a convolutional neural network (CNN) that is pre-trained using a set of color images are disclosed. The method comprises receiving a training dataset including multiple multidimensional images, each multidimensional image including a color image and a depth image; performing a fine-tuning of the pre-trained CNN using the depth image for each of the plurality of multidimensional images; obtaining a depth CNN based on the pre-trained CNN, wherein the depth CNN is associated with a first set of parameters; replicating the depth CNN to obtain a duplicate depth CNN being initialized with the first set of parameters; and obtaining a depth-enhanced color CNN based on the duplicate depth CNN being fine-tuned using the color image for each of the plurality of multidimensional images, wherein the depth-enhanced color CNN is associated with a second set of parameters.

Abstract: Abstract A method comprising: providing a first three-dimensional (3D) point cloud obtained according to a first sensing technique, a second 3D point cloud obtained according to a second sensing technique and a first radius of a sphere covering a real object underlying the second first 3D point cloud and a second radius of a sphere covering a real object underlying the second 3D point cloud; defining scales of the first 3D point cloud and the second 3D point cloud based on said first radius and second radius; searching statistically substantially similar candidate regions between the first 3D point cloud and the second 3D point cloud using an ensemble shape function (ESF); aligning the statistically substantially similar candidate regions between the first 3D point cloud and the second 3D point cloud at least in vertical direction; and ranking the aligned candidate regions between the 3D point clouds based on their structural similarity.

Abstract: A collation apparatus that collates an input image with registered images includes: a feature extracting unit extracting feature quantities from a region of an object in the input image; a calculating unit calculating, per each of the feature quantities, similarity degrees between the feature quantities extracted from the input image and corresponding feature quantities of the registered image, for each of the registered images; an integrating unit integrating, using a parameter, the similarity degrees for the feature quantities; an estimation unit estimating as to whether or not the object in the input image is identical with an object in each of the registered images; and a parameter learning unit learning the parameters to be used by the integrating unit, based on an estimation result of the estimation unit and the similarity degree integrated by the integrating unit.

Abstract: A method for represents the surroundings of a vehicle with at least two levels of resolution in a surroundings model, preferably for a vehicle having a sensor system for monitoring the surroundings and at least one driver assistance system. The surroundings model preferably represents a probability grid having a plurality of grid cells. Data from at least two grid cells at a higher level of resolution are combined and stored in a grid cell at a lower level of resolution.

Abstract: According to one embodiment, an inspection apparatus includes an optical system portion on which a plurality of reflected light rays corresponding to a plurality of observation directions with respect to an optical variable image device are incident and which images the plurality of reflected light rays incident thereon, an imaging portion that images the plurality of reflected light rays imaged by the optical system portion and thereby generates reflected light images corresponding the plurality of respective observation directions, an object relation determination portion that determines whether or not there is a predetermined geometric difference between objects that are contained in reflected light images imaged by the imaging portion and corresponding to different observation directions, and an authenticity determination portion that determines authenticity of the optical variable image device based on a result of the determination by the object relation determination portion.

Abstract: An image reconstruction apparatus and related method. The amount of out-field-of view material for a CT scanner (IMA) with a given field of view (FoV) in a bore (B) is established. Based on the measurement, a hybrid-image reconstructor (RECONX) is configured to switch between different reconstruction algorithms.

Abstract: According to an example method implemented by an imaging device, an image is recorded of a subject to which a wearable computing device is secured. Responsive to the recording, the imaging device wirelessly receives data from the wearable computing device. The data includes an identifier of the wearable computing device and biometric data of the subject. The identifier and biometric data are embedded as metadata in the recorded image.

Abstract: A method for detecting biometric characteristics in a captured biometric data image is provided that includes determining, by a processor, an approximate location for a biometric characteristic in a frame included in captured biometric data, and determining region of interest positions over the frame. Moreover, the method includes calculating a set of feature values for each position, generating a displacement for each set of feature values and generating a median displacement, and adjusting the biometric characteristic location by the median displacement.

Abstract: Some embodiments of the invention include a method for capturing and analyzing monitoring data of a measuring system. In some embodiments, the measuring system may include one or more sensors and being adapted for a measuring operation of a series of identical objects the measuring operation comprising a multitude of measuring sequences, each measuring sequence comprising the measuring of values of features of an object of the series, the method comprising a multitude of monitoring operations, wherein each monitoring operation comprises capturing monitoring data during a measuring sequence, the monitoring data of each measuring sequence including at least one image comprising the measuring system and/or a measurement environment, characterized by selecting a subset of measuring sequences from the multitude of measuring sequences; and visualizing an image sequence comprising the images of the monitoring data of the measuring sequences of the subset.

Abstract: A method includes: capturing an image of a coin, locating an acquisition area of the image using a landmark of the coin, the acquisition area comprising a feature unique to the coin and generating a digital representation of the acquisition area.

Abstract: An apparatus for detecting an afterimage candidate region includes: a comparison unit which compares gradation data of an n-th frame with integrated gradation data of an (n?1)-th frame and generates integrated gradation data of the n-th frame, where n is a natural number; a memory which provides the integrated gradation data of the (n?1)-th frame to the comparison unit and stores the integrated gradation data of the n-th frame; and an afterimage candidate region detection unit which detects an afterimage candidate region based on the integrated gradation data of the n-th frame, where each of the integrated gradation data of the n-th frame and the integrated gradation data of the (n?1)-th frame comprises a comparison region and a gradation region.

Abstract: A vehicle-in-question position detection unit (41) detects the position of the vehicle in question on the basis of GPS data acquired from a GPS receiver (3). An aerial photograph recording unit (43) records aerial photographs corresponding to the position of the vehicle in question. A peripheral object detection unit (42) produces a distance map on the basis of distance measurement data indicating the distance between a peripheral object and the vehicle in question obtained by a distance measurement sensor (2) mounted in the vehicle in question scanning the periphery. A peripheral map generation unit (44) updates past aerial photographs obtained from the aerial photograph recording unit (43) on the basis of the distance map, producing the latest peripheral map. The display device (5) displays the latest peripheral map at a screen.

Abstract: In an object detection apparatus 1, a window definer 11 defines a window relative to the location of a pixel in an input image 20. A classification value calculator 13 calculates a classification value indicative of the likelihood that a detection target is present in the window image contained in the window based on the feature data of the detection target. A classification image generator 14 arranges the classification value calculated from the window image according to the pixel location to generate a classification image. An integrator 15 integrates the classification image and a past classification image 42 generated from a past input image input prior to the input image 20 to generate an integrated image 45. A determiner 16 determines whether the detection target is present in the input image 20 based on the integrated image 45.

Abstract: A magnetic resonance (MR) system (10) generates an attenuation or density map. The system (10) includes a MR scanner (48) defining an examination volume (16) and at least one processor (54). The at least one processor (54) is programmed to control the MR scanner (48) to apply imaging sequences to the examination volume (16). In response to the imaging sequences, MR data sets of the examination volume (16) are received and analyzed to identify different tissue and/or material types found in pixels or voxels of the attenuation or density map. One or more tissue-specific and/or material-specific attenuation or density values are assigned to each pixel or voxel of the attenuation or density map based on the tissue and/or material type(s) identified as being in each pixel or voxel during the analysis of the MR data sets. In one embodiment, the tissue and/or material types are identified on the basis of a time series of MR phase images.

Abstract: A polyp detection apparatus and a method of operating the same are provided. The method of operating a polyp detection apparatus includes: generating an image of an object; detecting a polyp candidate region based on the generated image; detecting a reflected light region within the detected polyp candidate region; and displaying a final polyp region by excluding the detected reflected light region from the polyp candidate region.

Abstract: A determination method for determining reliability of a selective binding amount of a substance to be examined obtained as detection intensity of a label when a labeled substance to be examined binds to a selective binding substance fixed as a spot on a carrier includes: determining a position of the spot in image data obtained by imaging the detection intensity in the carrier and extracting a pixel group corresponding to the spot; calculating a ratio or a difference between a median value of the detection intensity of the pixel group extracted at the determining and a median value of the detection intensity of the pixel group excluding a certain top proportion of and/or a certain bottom proportion of pixels; and determining quality of the reliability based on the ratio or the difference calculated at the calculating and a certain reference value.

Abstract: An apparatus and a method for combining three-dimensional ultrasound images are provide. The method involves obtaining a plurality of three-dimensional ultrasound image data that corresponds to a Region of Interest (ROI); detecting one or more landmarks, using a parameter for detection; outputting the detection result and receiving a response from a user; registering each of one or more selected landmarks as link information according to a received response from the user; and generating a combined three-dimensional ultrasound image by combining at least two pieces of three-dimensional ultrasound image data using at least one of the one or more selected landmarks registered as the link information, wherein the at least two pieces of three-dimensional ultrasound image data commonly comprise the at least one of the one or more selected landmarks.

Abstract: A method, apparatus and computer program product are provided for determining spatial location for one or more facial features. A method computes features for an initial frame. The computed features of the initial frame generate a feature image. A method also determines whether a translation is verified between the initial frame and an intermediate frame, wherein a translation is verified in an instance in which a distance used to verify the translation between the initial frame and the intermediate frame is within a predetermined threshold level. A method also includes a face search, using a portion of the feature image, for one or more facial features, wherein the portion of the feature image searched is a fraction of the total number of frames analyzed in a feature computation cycle. A method also determines a spatial location for the one or more facial features detected in the intermediate frame.

Abstract: An image processing apparatus includes an image obtaining unit that obtains an image of a face, an edge enhancer that performs edge enhancement on the image and forms an edge-enhanced image, a binarizer that performs binarization on the edge-enhanced image and forms a binary image, and an area identifying unit that identifies an eyelash area in the image on the basis of the binary image.

Abstract: A gesture recognition apparatus capable of recognizing a hand by means of a binary image regardless of direction of light contacting the hand. A binarizing processing section binarizes an input image from a camera to a hand recognizing section to prepare a first binary image by a predetermined method. A rebinarizing processing section only rebinarizes a predetermined area of the input image to prepare a second binary image. A contraction processing section performs a contraction processing on the second binary image. The rebinarizing processing can increase the possibility of recognizing the hand by classifying a portion of the hand that was classified into black in the first binary image into white. The hand is determined to be recognizable if the hand can be recognized in the first binary image and/or in the second binary image.