Abstract: An image compression scheme is disclosed which models the human visual perception system. Using quantization of image error values, according to a visually-lossless scheme, an image can be compressed such that it is visually indistinguishable to the naked eye from the original image. To aid in image compression on portable devices such as a digital camera, the quantization can be precoupled into a look-up table.
Abstract: A method of generating image data comprising digital pixel data representative of one or more images on an input scanner support.In a first step the support is scanned to generate support image data at a first resolution comprising digital pixel data representative of the support and image.In a second step support image data at a second resolution is generated from the support image data at a first resolution. In a third step the boundary of an image is determined by performing a predetermined boundary determining algorithm on the support image data at the second resolution.In a fourth step, the image data is extracted from the support image data at the first resolution by extracting pixel data within the boundary of the image determined in the third step.
Abstract: In a decision-forest classifier in accordance with the invention, a decision forest including multiple decision trees is used to classify "seen" training data and "unseen" data. Each individual tree performs an initial classification based on randomly selected subsets of the data. The classification outcomes by the individual trees are combined using a discriminant process in the decision-forest classier to render the ultimate classification decision.
Abstract: Recently, a number of reversible wavelet transforms have been identified which allow for exact reconstruction in integer arithmetic. Different transforms vary in how rounding is performed. The present invention provides a transform, which is linear except for the rounding with non-linear operations in order to create a reversible implementation. Also, the present invention also provides transforms which are decomposed into all finite inpulse response parts.
Abstract: A color image display apparatus includes an input color characteristic obtaining unit for obtaining input color characteristic data simultaneously with input of image data, an input color characteristic data storage unit for storing the obtained input characteristic data, an image display device such as a CRT or the like, a display color characteristic data storage unit for storing the color characteristic of the image display device, a color transformation table generating means for generating a color transformation table, and a color transforming means for actually color-transforming the image data. The color transformation table generating means further includes a color transformation system selecting means, and a plural-tables generating unit.
Abstract: Flash correlation instantaneously compares two or more patterns to determine whether they are the same or essentially the same, regardless of the complexity of the images, and in spite of the addition of noise, local changes, and variations in resolution and focus. Flash correlation artifacts also provide cues to quantitative assessment of relative movement, stretching, blurring, and warping of one image with respect to the other. Presence of a flash correlation artifact between two stacks of patterns is sufficient to identify the stacks as containing at least one pair of massively correlated patterns. Precise control of registration, image size and orientation is not required. An image may be combined with other images, either by overlays or concatenation, and still be identified. A Flash correlation engine may utilize optical, analog or digital processing to provide rapid sorting, classification, and identification functions with minimal computational complexity.
Abstract: In order to progressively transmit a digital image (IM), a method for coding the image comprises the division (E1) of the image into vectors (V.sub.m,n), the selection (E2) of a code vector (VC.sub.d) associated with an index (A.sub.d) for each of the image vectors, and the replacement (E2) of the image vectors by the indices associated with the selected code vectors, in order to produce a plane of indices (PI) representing the image. The method comprises the steps of sub-sampling (E3) the plane of indices (PI) into sub-planes of indices (SPI.sub.1 to SPI.sub.F) respectively comprising indices selected from the plane of indices, and ordering (E4) the sub-planes of indices according to a predetermined order before transmitting them.
Abstract: An image signal having h.times.v pixel signals in the horizontal and vertical directions, respectively, is input to an image signal processor and converted into an image signal having m.times.n pixel signals in the horizontal and vertical directions, respectively. The converted image signal having m.times.n pixel signals is then converted into the image signal having h.times.v pixel signals and is then subjected to compression encoding. The compresed image signal is decoded and converted into the image signal having m.times.n pixel signals by thinning the pixel signals of the decoded image signal.
Abstract: An apparatus for vector-quantizing an image signal having a low transmission rate which includes a four-band linear filter para-unitary (LPPU) filter group for transforming an input image signal into transform coefficients which are represented by 4.times.4 frequency bands. A code modulator differential-pulse-code-modulates a direct current (DC) component among the transform coefficients output from the four-band LPPU filter group. A coefficient extractor extracts a predetermined number of low-frequency alternating current (AC) component coefficients from the transform coefficients output from the four-band LPPU filter group. A classifier determines a corresponding class, from a plurality of predetermined classes, using the low-frequency AC component coefficients extracted by the coefficient extractor, and outputs class information representing the determined class.
Abstract: A method, apparatus and computer program product are provided for querying by image colors using a JPEG format for images from the internet. First an image color is selected. A DCT coefficient is acquired from a JPEG image. The acquired DCT coefficient is compared with the selected image color. A match within the JPEG image is identified responsive to the acquired DCT coefficient being near the selected image color. An average color can be used for querying JPEG formatted images from the internet.
November 6, 1997
Date of Patent:
September 7, 1999
International Business Machines Corporation
Abstract: The present invention discloses a super resolution pan-sharpening technique that is used to increase the resolution of a multi-spectral signal using a panchromatic signal. Typically, multi-spectral signals, i.e., signals that contain a defined spectral band, have lower resolution than panchromatic signals since the multi-spectral signals normally contain fewer photons and require larger detectors for similar exposure periods. The technique of the present invention utilizes the panchromatic signal to increase the resolution of the multi-spectral signals to provide an increased resolution color output. The resolution of other spectral signals, such as infrared signals and UV signals, or any other desired signal, can also be increased in accordance with the techniques of the present invention.
Abstract: A mask area is specified as a target area of modification in a multi-tone image. The multi-tone image is converted to a plurality of binary images with a plurality of threshold values. A plurality of density-contours corresponding to the plurality of binary images are extracted. Without implementing any contraction or expansion, a portion overlapping the contour of the mask area is deleted from each extracted density-contour. A plurality of modified density-contours which do not include the portions overlapping the contour of the mask area are converted back to a modified multi-tone image.
Abstract: A fingerprint sensor includes a plurality of semiconductor devices adjacent a substrate and defining active circuit portions, and having only three metal layers. More particularly, the sensor may include a first metal layer interconnecting predetermined ones of the plurality of semiconductor devices; a second metal layer defining a ground plane; and a third metal layer comprising an array of electric field sensing electrodes connected to active circuit portions for generating an output related to a sensed fingerprint. The fingerprint sensor may also include a package surrounding the substrate and having an opening aligned with the sensing electrodes. In addition, a first external electrode may be carried by the package for contact by a finger. The sensor may thus also include an excitation drive circuit connected between the ground plane and the first external electrode for generating electric fields between the electric field sensing electrodes and adjacent finger portions.
May 16, 1997
Date of Patent:
August 17, 1999
Dale R. Setlak, Nicolaas W. Van Vonno, Rex Lowther, Dave Gebauer
Abstract: The present invention efficiently and quickly estimates signal (code value) dependant noise in an image and subsequently reduces that noise. The digital image is segmented both according to code value and into smooth vs. textured regions. Noise estimates taken from smooth regions are used to model noise as a function of code value. The predictions of the model are used to tone noise reduction in all areas of the image according to the appropriate code value.
March 24, 1997
Date of Patent:
July 13, 1999
Eastman Kodak Company
Patricia D Snyder, Thaddeus Francis Pawlicki, Roger Stephen Gaborski
Abstract: A digital copying apparatus provides high quality copy images suitable for an original document image while reducing a necessary capacity of a memory for storing image data. An image signal read by an image reading unit is binarized by a binary gradation processing unit in a gradation processing unit and is stored in a memory unit. Then, binary image data read from the memory unit is applied to a multi-value processing unit, wherein a pattern recognition unit determines whether or not the binary image data includes a pixel which requires a jagged line correction, and a halftone detector detects a halftone region in the binary image data. A determination unit and a selector select jagged line correction data output from a pattern memory and multi-valued image data converted from the binary image data by a multi-value filter. The selected data is output to an image forming unit.
Abstract: A color correcting device includes (1) a color correcting section furnished with a plurality of color correcting systems having their respective color correcting rates and color correcting accuracy, (2) a representative color selecting section for selecting representative colors from a color separation signal of an original image, and (3) a color correcting system selecting section for selecting one of the color correcting systems based on the color correcting result of the representative colors, whereby the color correcting device converts the color separation signal of the original image into a color correcting signal using the color correcting system selected by the color correcting system selecting section. This arrangement makes it possible to judge the quality of a color-corrected image without subjecting all the input image data of the original image to the color correcting operation. Consequently, a copy image of the original image can be made efficiently with satisfactory color correcting accuracy.
Abstract: A flat bed scanner of the type commonly used as a stand-alone computer peripheral in combination with a contact image sensor module, the glass cover of which is removed and the module placed under the glass surface of a flat bed scanner and which extends the full width of the scanning area to provide a 1:1 scanning ratio using a rod lens to focus the image onto the image sensors, with a monorail carriage for the scanner and including low friction strips to prevent tipping of the module into the glass, paper sensors to control the beginning and end of the scanning and for accomplishing scanning in both directions.
Abstract: Conventionally, when compressing data, since data (words) are sequentially retrieved from the head of a dictionary, a large amount of time is required for compression processing. Particularly when compression/expansion processing is performed a plurality of times, a large amount of processing time is required. In order to solve such problems, input data is dealt with as a set of words having a specific number of bytes, and a word retrieval table having a size equal to or larger than the size of a dictionary for registering words is provided. By storing the position of each word registered in the dictionary in the word retrieval table, it is unnecessary to sequentially retrieve data (a word) from the head of the dictionary, so that compression/expansion processing can be promptly performed.
Abstract: A method of estimating the chromaticity of illumination of a colored image consisting of a plurality of color-encoded pixels. The image colors are first mapped into an intensity-independent chromaticity space which is then divided into a plurality of separate regions. For each region, a first binary value is assigned to the region if the region contains no chromaticity value; or, a second binary value is assigned to the region if it does contain a chromaticity value. The assigned values are then applied as inputs to a pre-trained neural network having two output ports and at least one intermediate layer containing a plurality rality of ports connectible between selected input ports and the output ports. The chromaticity space values which characterize the input image's chromaticity of illumination are then derived at the output ports. The network is pretrained trained by initially connecting an arbitrary number of the intermediate layer ports to selected input layer ports.
November 15, 1996
Date of Patent:
May 25, 1999
Brian Vicent Funt, Vlad Constantin Cardei, Jacobus Joubert Barnard