Abstract: A method and apparatus for focusing an image in an image-capturing device utilizing at least one face in the image. The method includes detecting at least one face in the image, determining the size of the detected at least one face, determining the distance of the at least one face from the image-capturing device, wherein the determination utilizes the size of the detected at least one face, and focusing an image according to the determined distance of the detected at least one face.

Abstract: Automatic vision system object indexing and image database query system using both path-dependent and path-independent features of moving objects within a sequence of images. Feature vectors of both average over frames of an object traversing the field of view plus average over blocks of a grid for a path association. Color histograms may be an included feature.

Abstract: A digital still camera (10) receives and digitizes visible radiation (17) and sound waves (32) from a scene (12). When an operator actuates a shutter release (46), the camera detects and evaluates digitized information (22, 38) from the scene in a continuing manner, until a point in time when information representative of a human facial characteristic satisfies a specified criteria set by the operator through use of switches (47, 48) and a display (51). The camera then records in a memory (59) a digital image of the scene, which corresponds to the point in time. The stored image is a high-quality image, which can avoid characteristics such as eyes that are closed or a mouth that is open.

Abstract: A system (10) includes an image detector (12) which is a video camera, and which detects and outputs a succession of video images of a selected subject. A technique for processing these images involves preparing compressed image information. The compressed image information includes a first detected image (FIG. 2A) that serves as a reference image, includes information (FIG. 2G) identifying regions of subsequent detected images that differ from the first image, and excludes at least a portion of each of the subsequent images. The information identifying regions of subsequent detected images is free of contrast information for a substantial portion of each of those regions. The compressed image information for a given image is decompressed by displaying the reference image and by modifying the displayed reference image based on the information identifying differences from the reference image.

Abstract: A monitoring system (10) includes a video camera (12) which generates images of a monitored area. A computer (16) receives the images from the video camera, and contains a digitized map (85) of the monitored area. One of the detected images (84) is saved as a reference image. An operator defines a first region (86) corresponding to a selected portion of the monitored area, as viewed in the reference image, and defines a second region (87) which corresponds to the selected portion of the area as viewed on the map. Subsequent images from the camera are compared to the reference image, in order to identify an object of interest, and a first point associated with the object is identified. If the first point is within the first region, a warp transformation of the first point from the first region to the second region is carried out, in order to identify within the second region a second point which corresponds to the first point, and which identifies the location on the map of the object of interest.

Abstract: The method uses a three-dimensional face model and a technique called eigenface decomposition to analyze the video at one end. The facial feature locations and eigenface coding of the face image are sent to a decoder. The decoder synthesizes the face image at the receiving end. Eigenface decoding is used to texture map a three-dimensional model warped by detected feature locations.

Abstract: A method for deriving a two-dimensional first-range data model at a distance, d, from a two-dimensional second-range data model at a distance, d.sub.0. The method (10) comprises the steps of first smoothing the second-range data model, I, by masking the second-range data model by a masking matrix (step 24) and next undersampling the smoothed second-range data model (step 26) to yield a first-range data model. Both steps for smoothing (step 24) and undersampling (step 26) use operators depending on the ratio of the first-range distance, d, to the second-range distance, d.sub.0. One aspect of the invention includes a distance relating system (80) that performs the smoothing step (step 24) and undersampling step (step 26) to generate first-range data models from second-range data models in an intelligent sensor system.

Abstract: This invention is a method of decoding a stream of video image data transmitted as independent image frames consisting of plural marcoblocks transmitted in a nonsequential order. The method defines a sub-frame corresponding to a proper subset of the full frame. The method determines if a currently received macroblock is within the sub-frame. The method decodes the sub-frame. The sub-frame may be decoded at less than or equal to the frame rate of the video image data. A table has one entry for each macroblock that stores a transmission order within the video frame for the corresponding macroblock. The method determine if a current macroblock is within the sub-frame by reading the table. Each macroblock consists of a plurality of contiguous blocks and includes luminance data for any included blocks and chrominance data for the macroblock as a whole. The method optionally decodes the luminance data for each included block and ignores the chrominance data.

Abstract: A method for deriving a two-dimensional first-range data model at a distance, d, from a two-dimensional second-range data model at a distance, d.sub.0. The method (10) comprises the steps of first smoothing the second-range data model, I, by masking the second-range data model by a masking matrix (step 24) and next undersampling the smoothed second-range data model (step 26) to yield a first-range data model. Both steps for smoothing (step 24) and undersampling (step 26) use operators depending on the ratio of the first-range distance, d, to the second-range distance, d.sub.0. One aspect of the invention includes a distance relating system (80) that performs the smoothing step (step 24) and undersampling step (step 26) to generate first-range data models from second-range data models in an intelligent sensor system.

Abstract: A color resolution enhancement method and enhanced color cameras (600) with a single CCD detector (606) having attached color filters (604) and using luminance-based gradient interpolation to fill in color channels. A color camera, includes an image detector, the detector with sets S.sub.1, S.sub.2, . . . S.sub.N of pixels for N, an integer greater than 1, where pixels in the set S.sub.J detect radiation in a corresponding sampled channel C.sub.J for each J in the range 1 to N, a channel reconstructor coupled to an output of the detector. The reconstructor includes a luminance generator which combines the sampled channels C.sub.1, C.sub.2, . . . C.sub.N to a luminance channel, a gradient generator coupled to an output of the luminance generator and which transfers the luminance channel to a gradient channel, and a directional interpolator coupled to an output of the detector and to an output of the gradient generator and which transfers each of the sampled channels C.sub.1, C.sub.2, . . . C.sub.

Abstract: An image processor (10) is provided in which imaging data obtained by an image sensor (12) is processed for purposes of detecting motion. An optical flow field generator (16) processes the image data to provide condition number-based smoothed optical flow vectors, and stores those smoothed vectors in memory (18). The optical flow field stored in memory (18) is accessed and further processed by analyzer (20) for purposes of providing users with information on the motion of objects in the image plane of image sensor (12). Such user information may be displayed on a monitor (22).

Abstract: A method and apparatus for producing optical flow obtains a first image of a first spectral region of a field of view, obtains a second image of a second spectral region of the field of view, and processes the images to obtain the optical flow of the field of view. The first and second spectral regions are non-identical, as, for example, the visible and infrared bandwidths.

Abstract: An apparatus and method for improved segmentation and object recognition which describes objects and background appropriately based on calculations of desired local surface geometries such as local surface orientation, local surface curvature, surface extent and occluding boundaries, from sensed relative range images. Relative range data is used rather than converting such data to absolute range and then performing the desired geometric calculations.

Abstract: Operations of a plasma etch reactor (10) are monitored to detect aberrations in etching operations. A reference end point trace is defined (62) for the etch process. Regions are defined in the reference end point trace (70) with aid of a dynamic time warping matching function (84) and characteristics and tolerances for each region are defined (72-80). The etcher is run and an actual end point trace is obtained (82) from the running of the etcher. A warping function is constructed (88) between the actual trace and the reference trace. In building the warping function, candidate path segments (100) are constructed according to a minimum cumulative cost function (96). Once the regions of the reference trace and the actual trace has been matched according to an optimum dynamic time warping function path (106), characteristics of the matched regions are compared (66) to determine whether aberrations have occurred during the etch process.

Abstract: An improved apparatus and process for detecting aberrations in production process operations is provided. In one embodiment, operations of a plasma etch reactor (10) are monitored to detect aberrations in etching operations. A reference end-point trace (EPT) is defined (62) for the etch process. Regions are defined in the reference end-point trace (70) and characteristics and tolerances for each region are defined (72-80). The etcher is run and an actual EPT is obtained (82) from the running of the etcher. The actual EPT is analyzed to identify proposed regions of the actual EPT (86), and then the proposed regions of the actual EPT are matched with regions of the reference EPT (96). The system employs a series of heuristic functions in matching proposed regions of the actual EPT with regions of the reference EPT.

Abstract: A cursor control system for computer displays moves a cursor unambiguously in three dimensions using a two dimensional input device. The plane of movement of the two dimensional device is divided into logical regions which correspond to movement along a three dimensional axis. Movement of the two dimensional device into one of these regions causes the cursor to move along the corresponding axis of the display.

Abstract: An electronic system automatically detects significant events in a seismic trace. The seismic trace is defined as a series of consecutive points, equally spaced in time. The system provides for taking the Nth derivative of the trace at each of the points. The system has the capability of identifying the time of the zero-crossings of the Nth derivative and of establishing the amplitude of the trace at the time of the zero-crossing of the Nth derivative. Memory is provided for storing the amplitude of the trace and the times of the zero crossings in the Nth derivative. Finally, displays are provided for displaying the traces and the zero-crossings in the Nth derivative as a hard copy of visual display.

Abstract: A computer system reviews and modifies horizons tentatively defined by a network of pairs of turnings. The curvature for each of the paths of turnings of the network in both the X and Y directions is assessed. If the curvature exceeds a predetermined value in any of the paths, that path is terminated at the point of excessive curvature. Also, each pair of turnings is tested to determine whether it is in a closed loop made up of four turnings which in turn make up four pairs of turnings to determine that there is three-dimensional continuity. The horizons are thereby modified if modification is required.

Abstract: A computer system displays, in two dimensions, a representation of a three-dimensional volume. The three-dimensional volume representation is a tessellation of cells, represented as three-dimensional units. The three-dimensional volume is represented by a plurality of digital numbers. A designator designates which of the digital numbers are to be represented in the two-dimensional display. Each digital number is represented by one cell. Also, each digital number represents a color. Each corresponding cell is colored in accordance with the digital number. A designator designates which of the digital numbers is to be represented by a cell in the three-dimensional representation. A cursor, controlled by a mouse, may be used to indicate which of the three-dimensional volume representation is to be excavated by indicating at the appropriate cells and then electronically ordering the excavation.

Abstract: A computer system extracts horizons, each defined by a network of points in a first format from within a three-dimensional volume. Each point is tagged with a plurality of parameters including links in the X and Y direction for interconnecting the points. The system accesses any one of the points in the network then accesses every point that is linked directly or indirectly to the first accessed point. An address pointer is established for each of the accessed points, to define a second format. A second format is in a two-dimensional array and any point that is accessed a second time is rejected to prevent any spiraling of the horizon. The second format may be stored and displayed.