Abstract: In order to implement color-reduction processing more suitable, an image processing apparatus includes: a reference pixel determination unit configured to determine a reference pixel by scanning first image data; a color space division unit configured to set a plurality of regions in a hue circle and a plurality of luminance levels, and to determine a plurality of bins; a frequency distribution generation unit configured to generate a frequency distribution; a determination unit configured to select N bins based on the generated frequency distribution and to determine N types of pixel values respectively corresponding to the selected N bins; and a substitution unit configured to substitute each of pixels included in the first image data using the determined N types of pixel values.

Abstract: Aspects of the present invention are related to systems, methods and apparatus for image-based automatic defect detection.

Abstract: Methods, apparatuses and systems for storing approximations of multiple spatial pixel value adjustment surfaces for use in pixel value correction. The adjustment surfaces may be used for positional gain adjustment. A representation of the multiple adjustment surfaces is stored as coefficients representing a multi-dimensional solid of two-dimensional adjustment surfaces. The stored coefficients are used to determine a plurality of additional sets of coefficients which sets are then in turn used to determine pixel correction values of adjustment surfaces.

Abstract: Performing real-time compression on an image for target buffer fullness includes dividing the image into N macro-blocks, performing a discrete cosine transformation (DCT) on each of the N macro-blocks, defining a Quantization Parameter Scalar (Q) for each of the N macro-blocks of the image on the DCT being performed, initializing the Quantization Parameter Scalar (Q) for the first Macro-block to a value that correlates to a buffer fullness of a previously compressed image, and monitoring the buffer fullness by comparing the buffer fullness with the target buffer fullness. The N macro-blocks include 16×16 macro-blocks. The Q value is increased to a first new value when the buffer fullness is greater than the target buffer fullness. The Q value is decreased to a second new value when the buffer fullness is less than the target buffer fullness.

Abstract: An image processing device that processes multivalue image data includes: a histogram storage section that stores an appearance frequency of each of gradation values; a palette storage section that stores the gradation value that corresponds to each of index values; an output section that accesses the histogram storage section data and outputs the appearance frequency of the gradation value of the piece of pixel data; a histogram generator that accesses the histogram storage section for each piece of the pixel data included in the image data and adds one to the appearance frequency of the gradation value of the piece of pixel data; and a palette generator that assigns, when the appearance frequency that is output from the output section indicates 0, the index value to the gradation value and accesses the palette storage section and stores the gradation value that corresponds to the index value.

Abstract: An image processing apparatus, which performs an error diffusion on M-value image data with an error weight matrix into N-value image data, includes: a corrected-value calculating unit that calculates a corrected value by adding a first value and a second value to a pixel value corresponding to a processing-target color component of a pixel of interest, the first value being obtained by multiplying an error for each quantized pixel of the processing-target color component by the matrix, and the second value being obtained by multiplying an error for each quantized pixel of a different color component by the matrix; a quantizing unit that compares the corrected value with a quantization threshold, and outputs N-value image data; and an error diffusion unit that calculates an error at the pixel of interest by subtracting a quantization threshold from a value obtained by adding the first value to the pixel value.

Abstract: Color management converting source-side color images into destination-side color images. A module library includes module entries corresponding to function modules implementing color processing functionalities. Each module entry includes a module locator, and some include a profile interface definition. A profile library includes profile entries corresponding to sources of parameters. Profile entries include a profile locator and a profile interface definition. A rule library includes external and internal rules which characterize color transformation workflows. Each external rule is associated with function modules. An interface receives factual input including factual input derived from source-side color image data. A rule engine determines a sequence of function modules and sources of parameters by using the factual input and the plural and external rules in the rule library, and builds the color transformation workflow from the determined sequence of function modules and sources of parameters.

Abstract: A method and apparatus provides for block based image compression with multiple non-uniform block encodings. In one embodiment, an image is divided into blocks of pixels. In one embodiment the blocks are four pixels by four pixels, but other block sizes are used in other embodiments. In one embodiment, a block of pixels in the original image is compressed using two different methods to produce a first and second compressed block. Thus, each block in the original image is represented by two, typically different, compressed blocks. In one embodiment, color associated with a pixel is determined by combining the compressed information about the pixel in the first compressed block with information about the pixel in the second compressed block. In another embodiment, global information about the image is combined with the information in the first and second compressed blocks.

Abstract: A method implemented in a graphics engine for decoding image blocks to derive an original image is provided. The method comprises receiving at least one encoded image data block at a block decoder, the at least one encoded image data block comprising a plurality of codewords and a bitmap. The method further comprises determining a block type based on the plurality of codewords and selecting a decoder unit among a plurality of decoder units in accordance with the block type.

Abstract: A system and a method for document image segmentation have been disclosed. Image segments are obtained by forming different clusters in a document image. The document image may include images of company logos, product marks or trademarks. The invention can perform image segmentation on any kind of complex colored image and can recognize logos, product-marks or trademarks which comprise text or graphics, wherein the text can be either of uniform font style or uneven font style such as fancy font styles, calligraphic styles or having different orientation.

Abstract: A system and method for performing multi-image training for pattern recognition and registration is provided. A machine vision system first obtains N training images of the scene. Each of the N images is used as a baseline image and the N?1 images are registered to the baseline. Features that represent a set of corresponding image features are added to the model. The feature to be added to the model may comprise an average of the features from each of the images in which the feature appears. The process continues until every feature that meets a threshold requirement is accounted for. The model that results from the present invention represents those stable features that are found in at least the threshold number of the N training images. The model may then be used to train an alignment/inspection tool with the set of features.

Abstract: A frame rate improving apparatus and method use a motion trajectory. The frame rate improving apparatus determines forward feature point trajectory information through a feature point, and generates a new intermediate frame by performing backward motion estimation through the feature point trajectory information, thereby generating an interpolated image with high quality.

Abstract: A document to be segmented is converted into a common representation format, if necessary. Parsing of the document results in a document model that is analyzed based on at least one structure-dependent function to identify segments within the document. In one embodiment, the structure-dependent function may comprise a template, or a best-fit template of a plurality of templates, used for comparison with the document model. In other embodiments, the structure-dependent function may comprise table of contents information, font properties within the document model and/or an average segment size determined according to previously identified segments in one or more additional documents that are related to the document under consideration. Semantic-content dependent functions may be applied to further refine the analysis by identifying sub-segments within the extracted segments, or by identifying segments that may be properly merged according to the similarity of their respective semantic content.

Abstract: A Two-Dimensional (2D) image segmentation apparatus for segmenting pixels of a progressive input 2D image includes a group information storing unit storing information of pixel groups including a plurality of adjacent pixels; a pixel determining unit determining coordinates of an input pixel, and determining whether the input pixel is an effective pixel for segmentation; a group scanning unit scanning a adjacent pixel group disposed in a scan range preset from the effective pixel in the group storing unit when the input pixel is determined as the effective pixel by the pixel determining unit; and a group information updating unit updating information of a pixel group stored in the group information storing unit according to whether there is the scanned adjacent pixel group and a pixel group including an input pixel preceding the effective pixel. Moreover, a method is provided that removes red-eye using area information of a pixel group determined by the segmentation apparatus.

Abstract: A technique for optimizing a prediction method of samples in blocks of an image is provided. The image includes a first block, a second block, a third block, and a fourth block, each of the blocks include 8×8 blocks and form one Macro block. The method includes performing a prediction of the second block, the third block and the fourth block by performing at least one of prediction methods. A prediction error per block (Pe) is computed for each prediction method that is performed to predict the second block, the third block, and the fourth block. The prediction error per block (Pe) equals an original block value minus a predicted block value. An optimal prediction method is chosen from each of the prediction methods performed that results in minimum ?|Pe| pixels per block (summation on pixels per block). The optimal prediction method and the Pe for each block are stored.

Abstract: A method of pre-processing an image to identify processes for subsequent processing of the image, comprising the steps of: a) investigating portions of the image using a spatial filter; b) calculating for a first plurality of regions within a portion of the image under investigation respective metrics as a function of intensity within those regions; c) selecting combinations of regions within the portion of the image under investigation and processing them to obtain a second plurality of filter values, where the second plurality is greater than the first plurality; and d) comparing the filter values with process thresholds for subsequent processes so as to identify subsequent processes that can be skipped.

Abstract: An image processor includes a hue region judging section judging which one of a plurality of hue regions corresponds to a hue and a converting section obtaining a correction value by using the hue to correct the hue based on the correction value. The converting section obtains a first difference value by using a difference between the hue and a first reference value, obtains a second difference value by using a difference between a second reference value and the first difference value, and obtains the correction value by using the second difference value. The hue is corrected based on the correction value.

Abstract: A recording device includes: an encoding unit configured to perform compression encoding on single image data obtained by imaging a subject from a predetermined viewpoint to generate first compression-encoded data, and perform compression encoding on composite image data in which a plurality of image data obtained by imaging the subject from mutually different viewpoints have been composited, to generate second compression-encoded data; a file generating unit configured to generate a first image file including the first compression-encoded data generated at the encoding unit, and generate a second image file including the second compression-encoded data generated at the encoding unit; and a file writing unit configured to write the first image file and the second image file generated at the file generating unit, to a recording medium in a correlated manner.

Abstract: A method for segmenting a digital image into a plurality of target objects, comprising, generating a plurality of probability maps of the image, wherein each probability map is derived from a different segmentation classifier; generating a combined probability map based on the plurality of probability maps; mapping a plurality of image points based on one or more local object maxima; applying one or more object constraints based at least in part on the mapped points to identify local object information; applying one or more regional thresholds to the combined probability map, given the local object information and a background mask, to segment the image into regions; creating a segmented image at least in part by merging the segmented regions with corresponding local object maxima; and at least temporarily storing or displaying the segmented image on a digital device.

Abstract: An image processing apparatus includes a storage device, an image detection circuit, a compression circuit, a decompression circuit, and an overdrive processing circuit. The image detection circuit generates a compression mode control signal according to a first frame. The compression circuit compresses an image data of a second frame according to the compression mode control signal, thereby generating a compressed image data of the second frame to the storage device. The first frame precedes the second frame. The decompression circuit decompresses the compressed image data of the second frame read from the storage device according to the compression mode control signal, thereby generating a recovered image data of the second frame. The overdrive processing circuit determines overdrive voltages of a third frame according to an image data of the third frame and the recovered image data of the second frame, where the second frame precedes the third frame.