Abstract: A unified system and method for encoding an array of data. If the data array is comprised of DCT data, then coefficients from corresponding positions in the data array are mapped into a common blocks in a second data array so as to group similarly valued coefficients. If the data array is comprised of wavelet data and the wavelet tile is greater than a predetermined size, then the wavelet tile coefficients are mapped into a second array so as to combine coefficients from the same wavelet family. After the DCT or wavelet coefficients have been mapped, the DC coefficients are encoded using a differential pulse code modulation (DPCM) process. The maximum number of bits required to represent any coefficient in each block family in the data array is determined. The difference between the maximum number of bits required to represent any coefficient in the entire data array and each of the block family maximums is determined and encoded.

Abstract: A system, method, and software product provide for unsupervised identification of complex, nonlinear subspaces in high dimensional data. The system includes a vector quantization module, a weighted topology representing graph module, and an encoding module. The vector quantization module takes vector data inputs and extracts a group of inputs about a number of cluster centers, using a globally optimized clustering process. The weighted topology representing graph module creates a weighted graph of the vector space, using the cluster centers as nodes, weighting edges between nodes as a function of the density of the vectors between the linked nodes. The encoding module uses the weighted graph to recode the input vectors based on their proximity to the cluster centers and the connectedness of the graph. The recoded vectors are reinput into the vector quantization module, and the process repeated until termination, for example at a limited number of cluster centers.

Abstract: An inventive method tracks IC devices through the assembly steps in a manufacturing process. Prior to die attach, a laser scribe marks the lead frame of each of the devices with a coded hole matrix that gives each device a unique ID code. During die attach, an optical hole reader retrieves the ID code of each of the IC devices, and a computer system stores the retrieved ID codes in association with the lot numbers of the ICs attached to the lead frames. The ID codes of the devices are then read at each step in assembly so the devices can be tracked through assembly individually, rather than by lots. As a result, the devices can proceed through assembly in a more efficient, continuous manner (i.e., without breaks between lots).

Abstract: A process for image analysis which includes selecting a number M of images, forming a model graph from each of the number of images, such that each model has a number N of nodes, assembling the model graphs into a gallery, and mapping the gallery of model graphs into an associated bunch graph by using average distance vectors &Dgr;ij for the model graphs as edge vectors in the associated bunch graph. A number M of jets is associated with each node of the associated bunch graph, and at least one jet is labeled with an attribute characteristic of one of the number of images. An elastic graph matching procedure is performed wherein the graph similarity function is replaced by a bunch-similarity function.

Abstract: A computer-based method and system for digital 3-dimensional imaging of an object which allows for viewing images of the object from arbitrary vantage points. The system, referred to as the Lumigraph system, collects a complete appearance of either a synthetic or real object (or a scene), stores a representation of the appearance, and uses the representation to render images of the object from any vantage point. The appearance of an object is a collection of light rays that emanate from the object in all directions. The system stores the representation of the appearance as a set of coefficients of a 4-dimensional function, referred to as the Lumigraph function. From the Lumigraph function with these coefficients, the Lumigraph system can generate 2-dimensional images of the object from any vantage point. The Lumigraph system generates an image by evaluating the Lumigraph function to identify the intensity values of light rays that would emanate from the object to form the image.

Abstract: A method and apparatus for decoding, e.g., an MPEG-like video information stream comprising quantized discrete cosine transform (DCT) coefficients representing relatively high resolution image information to produce corresponding pixel blocks having a relatively low image by modifying a quantization matrix to dequantize DCT coefficients in a manner that allows for reduced complexity transformation of the dequantized DCT coefficients into sub-sampled image domain information.

Abstract: An ultrasonic imaging system includes an ultrasonic transducer having an image data array and a tracking array at each end of the image data array. The tracking arrays are oriented transversely to the image data array. Images from the image data array are used to reconstruct a three-dimensional representation of the target. The relative movement between respective frames of the image data is automatically estimated by a motion estimator, based on frames of data from the tracking arrays. As the transducer is rotated about the azimuthal axis of the image data array, features of the target remain within the image planes of the tracking arrays. Movements of these features in the image planes of the tracking arrays are used to estimate motion as required for the three-dimensional reconstruction. Similar techniques estimate motion within the plane of an image to create an extended field of view.

Abstract: A method of generating an upsampled target pixel positioned between two lines of input source data includes the step of comparing pixels of different lines of the source data in a region surrounding the upsampled target pixel to be generated in at least two different directions. An interpolation direction based on the comparison is selected and interpolations between selected pixels of the source data in the determined interpolation direction are carried out to compute intermediate pixels on a line segment passing through the upsampled target pixel. An interpolation between the intermediate pixels is carried out to generate the upsampled target pixel. An apparatus for performing the method is also disclosed.

Abstract: A novel technique for decoding variable length encoded data in a bit stream such as video data compressed in accordance with JPEG, MPEG, or DV standards is disclosed. The disclosed technique preprocesses the input data to generate a pointer to a variable length code table from the encoded code word, the entry in the variable length table providing the information necessary to decode the code word. Preprocessing of the input data includes encoding leading ones or leading zeros and selecting a subset of bits following the leading ones or zeros. The pointer to the variable length code table is generated from a base value corresponding to the symbol type being decoded. The base values for the symbols being decoded may be stored in a first look up table which may contain multiple entries for one or more of the symbol types, the number of entries corresponding to the number of leading ones or zeros.

Abstract: A real-time hotspot detection system is disclosed. In a preferred embodiment, the real-time hotspot detection system of the present invention comprises a photolithography stepper that has been modified to output z-height focus data for each cell to a fault detection module of the present invention. After the entire wafer has been exposed by the stepper, the fault detection module computes the mean and standard deviation and/or other statistical data for the wafer data and compares the computed standard deviation with a predetermined limits for “normal” standard deviation, representative of a completely flat wafer having no hotspots thereon. If the computed standard deviation is outside the limits for the normal standard deviation, the fault detection module provides control signals to the stepper for halting the operation thereof and for generating an indication to the stepper operator, via a pole light associated with the stepper, that a failure has occurred.

Abstract: Methods and systems for data compression and decompression are disclosed. Specifically, the method for data compression of the present invention receives four inputs representing pixel information of an image and performs a combination of a 4-point discrete cosine transform (DCT) and a weighting function to obtain four outputs representing digital representations of an image. The method for data decompression receives four inputs representing digital representations of an image and performs a combination of an inverse weighting function and a 4-point inverse discrete cosine transform (IDCT) to obtain pixel components of an image. The methods and systems of the present invention may be used in optimizing digital video encoders and decoders.

Abstract: Brightness values of picture elements of an image segment are subjected to a conformal imaging, so that the brightness values of the picture elements that are located at an edge of the image segment are imaged onto picture elements of an edge of a target image segment having a prescribable shape. Interpolation of the brightness values is undertaken between the imaged brightness values of the target image segment.

Abstract: A method and an arrangement are presented for pattern recognition on the basis of statistics. According to the method, for an object to be recognized on the basis of a complete set of two-class or multiclass classifiers, the association with each target class of the class set is estimated with a numerical value that is produced by cascaded use of polynomial classifiers. According to the invention, on a learning sample in which all class patterns to be recognized are sufficiently represented, there is a selection, from all the two-class or multiclass classifiers by way of their estimation vector spectrum, of those two-class or multiclass classifiers with estimations contributing the most to minimize a scalar quantity calculated over the estimation vector spectrum and having high separating relevance. The selected two-class or multiclass classifiers are subsequently used to form, via an expanded learning sample, estimation vectors from which expanded characteristic vectors are produced by polynomial linking.

Abstract: An apparatus for determining the profile of an object provides a sensor having at least one light source which projects a sheet of light that is intersected by the object, whereby a stripe of points on the object are illuminated. The light source projects the sheet from multiple angles, so that the resulting stripe represents a continuous profile. The profile is collectively viewed by at least two optical detectors, which each have a different view than the others and provide an output signal representing its particular view of the profile. The output signals are received by an image processor, which digitizes the signals, provides filtering, and identifies (extracts) the points in each view that represent the profile. The image processor uses a predetermined set of calibration parameters to transform the coordinates of the extracted points to corresponding coordinates in a common reference system that collectively represent the profile of the object.

Abstract: A method and apparatus for color calibrating an imager device is disclosed. The imager device is subjected to a plurality of light sources. Color channel responses are obtained from the imager device and the color calibrating coefficients are determined.

Abstract: The present invention provides an image collation device having a simple whole structure which is applied, for example, as a fingerprint collation device. Linear images cut from the one image is stored in a memory means, and the other image is collated with the stored liner images. Thereby the image data of the linear images can be recorded in a memory means having a small capacity.

Abstract: A method for generating a stream of image data which is to be transmitted via a data channel, in particular a mobile radio channel. In this method, the images are divided into a quantity of macroblocks. In addition, the image information elements are divided into classes of information elements of the same kind. The image data of at least a portion of the classes are inserted into the stream by macroblocks. In a deviation from the proposed H.263 standard, the image data in each macroblock which have a fixed word length, in particular the DQUANT and INTRADC values, are combined into a class (ADMIN+DC-INTRA) provided for them, and are inserted consecutively into the stream with no variable-length image data being transmitted between them. A method for analyzing a stream of data is also provided.

Abstract: The system, method, program, and print pattern of this invention allows print misregistration to be detected and controlled through density measurements. A special print pattern is used for which a correlation between a density measurement of the print pattern and an amount of misregistration can be made. Initially, for a given printer, and for each printing station within the printer, the special print pattern is printed using two printing stations at varying amounts of misregistration of one of the printing stations. The density of each special print pattern printed at the varied amount of misregistration is measured. A correlation is made between density and amount of misregistration of the one printing station relative to the other printing station. Whenever misregistration is desired to be determined, the special print pattern is printed, the density is measured, and the correlation is applied to the measured density to determine a misregistration amount.

Abstract: A system and method for generating bounded-loss color transformations employed in the compression and decompression of multi-spectral images is provided. The bounded-loss color space transformations provide transformations from a first color space to a second color space and back to the first color space with bounded loss or error. The bounded loss or error allows for the near lossless compression and reconstruction of multi-spectral images transformed from a first color space to a second color space and ultimately back to the first color space.

Abstract: The background grey-level of a document is determined based on the gain of the document. A histogram is generated and compressed. The standard deviation of the distribution curve of the compressed histogram is determined. A gain factor is determined using the mean and standard deviation. Using the background grey-level, the dynamic range of the document is adjusted.