Abstract: An image processing apparatus includes: image acquiring means for acquiring an image; search-window-size setting means for setting a size of a search window; search-range setting means for setting a search range in the image in relation to the set size of the search window; scanning means for moving the search window having the set size in the set search range; face-area determination means for determining whether the image in the search window at each scanning position is a face area; and face-information output means for outputting information of the face area obtained from a determination result of the face-area determination means.

Abstract: Techniques for positioning images in an electronic document. The images may encode information that can be decoded by devices configured to read the images. An area within the electronic document for positioning the image and a set of one or more colors to be used for the image are determined. The determined area and set of colors are such that the image when positioned in the electronic document blends in with the contents of the electronic document while still being clearly visible for reading by reader devices that are configured read and decode the information encoded by the image.

Abstract: A system for performing a three dimensional stereo measurement that uses a sensor for obtaining a sensor image of a scene, and a database for providing first and second reference images of the scene that are a stereo pair of images. At least one processing system is responsive to an output of the sensor and in communication with the database. The processing system registers the sensor image with the first reference image, and also selects a point of interest from one of the sensor image and the first reference image. The processing system performs a stereo point measurement from the selected point of interest and the first reference image to determine a point in the second reference image that represents a stereo mate of the selected point in the first reference image.

Abstract: There is a need to provide simple, accurate, fast and computationally inexpensive methods of object and hand pose recognition for many applications. For example, to enable a user to make use of his or her hands to drive an application either displayed on a tablet screen or projected onto a table top. There is also a need to be able to discriminate accurately between events when a user's hand or digit touches such a display from events when a user's hand or digit hovers just above that display. A random decision forest is trained to enable recognition of hand poses and objects and optionally also whether those hand poses are touching or not touching a display surface. The random decision forest uses image features such as appearance, shape and optionally stereo image features. In some cases, the training process is cost aware. The resulting recognition system is operable in real-time.

Abstract: To accurately discriminate a dot region within an image, the image processing apparatus includes first through fourth isolated dot discriminating portions each judging, for each of a plurality of pixels included in the image, whether the pixel corresponds to a center pixel of an isolated dot, an isolated dot size determining portion detecting an isolated dot size, a dot region discriminating portion judging whether a target pixel is included in a dot region based on a position of the pixel judged as being the center pixel of the isolated dot, and a dot region determining portion determining the dot region based on a position of the target pixel judged as being included in the dot region and the detected isolated dot size.

Abstract: An apparatus for picking up objects including a robot for picking up an object, at least one part of the object having a curved shape, having a storing means for storing a gray gradient distribution model of the object, a recognizing means for recognizing a gray image of the object, a gradient extracting means for extracting a gray gradient distribution from the gray image recognized by the recognizing means, an object detecting means for detecting a position or position posture of the object in the gray image in accordance with the gray gradient distribution extracted by the gradient extracting means and the gray gradient distribution model stored by the storing means, a detection information converting means for converting information of the position or position posture detected by the object detecting means into information of position or position posture in a coordinate system regarding the robot; and a robot moving means for moving the robot to the position or position posture converted by the detection

Abstract: A biometrical feature inputting apparatus includes a 1-dimensional or quasi 1-dimensional image sensor. When a finger and the image sensor are relatively slid, a finger sliding guide keeps a finger and an effective pixel unit of the image sensor to a constant distance without any contact between them. An image processing section sequentially generates partial images by imaging emission light that is scattered inside the finger and then emitted from a skin surface of the finger by the image sensor during the relative motion of the finger and the image sensor and link the partial images to an image.

Abstract: Change image detecting devices and methods are disclosed. In one example, a change image detecting device includes a change image detecting unit that selects a reference image and an image to be processed and a determination unit that compares an Na-th (‘Na’ is a natural number equal to or smaller than N) partial region of N (‘N’ is an integer equal to or larger than 2) partial regions. If the determination unit determines that there is a change, the change image detecting unit selects the image to be processed or an image, which is temporally later than the image to be processed, as a new reference image and detects the image to be processed or the image, which is temporally later than the image to be processed, as the change image.

Abstract: A finger sensing device may include a finger sensing area to receive a user's finger moved in a sliding motion, and a controller cooperating with the finger sensing area for generating successive image samples. Moreover, the controller may also generate the displacement estimate of the user's finger by at least performing a plurality of different image sample correlations between at least one pair of image samples, and cross-verifying results of the plurality of different image sample correlations.

Abstract: A scanning system for scanning a wire to determine the characters provided on the wire.

Abstract: A method and apparatus for correcting misalignment between fields of view of a CT device and a NM device of a modular multimodality medical imaging system, by providing a Field Of View Calibration Matrix (FOV-CM) containing rotational and translational transformations between coordinate systems of the CT and NM systems.

Abstract: A scanning system for scanning a wire to determine the characters provided on the wire.

Abstract: The invention relates to a method and system for the acquisition and correlation matching of points belonging to a stereoscopic pair of images, whereby the pair is formed by a first image and a second image representing a scene. According to the invention, the two images of the pair are acquired with a single acquisition instrument (30) comprising two sensors CCD (31, 32) in the optical focal plane. The matching of the acquired stereoscopic pair consists in determining, by means of correlation, the point in the second image that is homologous to a point in the first image. Said correlation is performed for a point from the first image using an optimally-sized correlation window. When the homologous point of a point from the first image has been determined, the position deviation between the point from the first image and the homologous point thereof is entered in a table. Once all of the homologous points of the points from the first image have been found, the results table is reset barycentrically.

Abstract: A depth-based image enhancement system is disclosed. A depth estimation unit generates three-dimensional (3D) depth information from a two-dimensional (2D) image. Subsequently, an image enhancement unit enhances the 2D image according to the 3D depth information.

Abstract: Based on an area detection signal, a layer separation section outputs a text component of a document, to a feature point calculating section, and generates four layers from a pictorial component of the document to output the generated layers to the feature point calculating section. The feature point calculating section sums feature points extracted for each component. A features calculating section calculates a hash value based on the feature points. A vote processing section searches a hash table based on the hash value, and votes for a reference image associated with the hash value. Based on the voting result, a similarity determination processing section determines whether the document image is similar to any reference image, and then outputs the determination result. Thus, even if the document contains a photograph, accurate matching can be performed.

Abstract: A system of generating three-dimensional (3D) depth information is disclosed. A color and object independent local blurriness estimation unit analyzes blurriness of each pixel of a two-dimensional (2D) image. Subsequently, a depth assignment unit assigns depth information to the 2D image according to the analyzed blurriness.

Abstract: An acquisition system for obtaining images of a fingerprint from a finger, or marks of another kind of target. The system may have two or more cameras positioned at different directions towards the target. Each camera may have a light source for illuminating the target from a direction different than that of the respective camera. The cameras may take sequences of images of the target at different focuses as the whole target might not be in focus in one image due to a depth of focus being less than the depth of the target. Portions of the images showing the target in focus may be cut from the images and stitched together to result in an image revealing virtually the whole target in focus. This target may be a fingerprint to be rolled out on or in a two dimensional medium for analysis, identification, storage, and so on.

Abstract: A method for performing a convergence calculation using a projective transformation between images captured by two cameras to observe a flat part of an object in the images, wherein a computational load is reduced while securing a convergence property of the convergence calculation. Initial values (n0(i), d0(i)) are set to values satisfying a limiting condition that should be satisfied by the initial values (n0(i), d0(i)), where the limiting condition is that a plane ?a(i) defined by the initial values (n0(i), d0(i)) of given types of parameters (n(i), d(i)) of a projective transformation matrix in the convergence calculation is inclined with respect to an actual plane including the flat part of the object to be observed.

Abstract: An image processing device includes a deformation processing unit that deforms at least a portion of an area on a target image, which is generated by photographing, based on a distance between a photographic device and a subject of the target image. A deformation amount adjustment unit uses a distance parameter that correlates with the distance between the photographic device and the subject of the target image when photographing to thereby adjust a degree of deformation to increase as a distance indicated by the distance parameter becomes shorter.

Abstract: The method, system, and apparatus of source statistics based intra prediction type is disclosed. In one embodiment, a method includes classifying a four-pixel square block in an edge class (e.g., may include a DC edge class, a vertical edge class, a horizontal edge class, a diagonal edge class, and/or a planar edge class) based on an edge classifier, classifying an eight-pixel square block having the four-pixel square block and other four-pixel square blocks as a homogenous class if the four-pixel square block and the other four-pixel square blocks of the eight-pixel square block belong to the edge class, assigning a direction to the edge class of the eight-pixel square block, and determining an optimal intra-prediction type through the classification such that empirical testing of all possible ones of the edge class and the direction is avoided when the homogenous class is identified.