Abstract: A system, method, and computer program product are provided for automatically segmenting input document images into regions of black text, white space, and image content. A set of scanned training documents representing the range of text and images to be processed is coarsely tagged to classify regions by content type. The training images are divided into bricks, parameters describing individual brick features are evaluated, and the bricks are classified according to the parameter values. A classification map that relates parameter values to classification codes describing content type is constructed by generating linear equations separating a parameter space into parameter regions along classification boundaries. After training, input documents are scanned and divided into bricks, and brick parameters are converted into an index into the classification map, to classify document regions by content.

Abstract: An apparatus and a method are described for processing and reproducing a document including both text and half-tone images without generating undesirable image artifacts. A band-pass FIR filter with a predetermined frequency response is used to transform pixels in an input image to sharpen text edge and smooth pre-half-toned areas, and perform error diffusion and cluster dot screening methods on the transformed pixels. The FIR filter has unity gain at DC to preserve image brightness. An integrated error diffused cluster-dot screening approach is used to suppress moiré and other image artifacts, while providing a robust, crisp output.

Abstract: A system, method, and computer program product are provided for automatically segmenting input document images into regions of black text, white space, and image content. A set of scanned training documents representing the range of text and images to be processed is coarsely tagged to classify regions by content type. The training images are divided into bricks, parameters describing individual brick features are evaluated, and the bricks are classified according to the parameter values. A classification map that relates parameter values to classification codes describing content type is constructed by generating linear equations separating a parameter space into parameter regions along classification boundaries. After training, input documents are scanned and divided into bricks, and brick parameters are converted into an index into the classification map, to classify document regions by content.

Abstract: An apparatus and a method are described for processing and reproducing a document including both text and half-tone images without generating undesirable image artifacts. A band-pass FIR filter with a predetermined frequency response is used to transform pixels in an input image to sharpen text edge and smooth pre-half-toned areas, and perform error diffusion and cluster dot screening methods on the transformed pixels. The FIR filter has unity gain at DC to preserve image brightness. An integrated error diffused cluster-dot screening approach is used to suppress moiré and other image artifacts, while providing a robust, crisp output.

Abstract: A stochastic screening mask is provided for rendering halftone images. A filter controlled force masking method is used to vary the mask patterns from regularity to randomness. Dot size (amplitude), dot number (frequency), and dot shape are modulated integrally at each gray level during the generation of the mask. The single pixel limit for blue noise masking is broken down by this approach. The halftone patterns generated with this method can have visually pleasing blue noise attribute. This new screening method would provide flexible solutions for different types of printing processes.

Abstract: The present invention is directed to a method of discriminating between textual content and graphical content. The method includes the steps of receiving a plurality of pixel values for a pixel line segment, calculating a plurality of spatial gradients based on the pixel values of adjacent pixels, determining a smoothness index by processing the plurality of spatial gradients, and identifying the pixel line segment as text or graphics by comparing the smoothness index to a threshold value.

Abstract: A system-on-a-chip controller having a first processor and a second processor. The first processor provides control processing and image processing. The second processor provides image processing. The processors receive data from an external source through a data bus. Also, the controller can include a third controller to provide I/O functionality to an external device. The second processor processes the stored data in either a row or column configuration. A fixed-length instruction word can be decoded into two instructions, an operation instruction and an I/O instruction, and can be used to process the data. The I/O instruction can be disposed in an unused bit field of the operation instruction.

Abstract: A signal conversion system parses a multi-dimensional input signal into most-significant and least-significant portions. The most-significant portions are used as an index to a look-up-table. Stored in each entry of the look-up-table is a base value and multi-dimensional differential values. The system generates an output representation of the signal by combining the corresponding base value with interpolated differences determined by interpolation performed using the differential values and the least-significant portions. Interpolation error is decreased by storing intermediate differential values. Output processing is used for linearization.

Abstract: A system-on-a-chip controller having a first processor and a second processor. The first processor provides control processing and image processing. The second processor provides image processing. The processors receive data from an external source through a data bus. Also, the controller can include a third controller to provide I/O functionality to an external device. The second processor processes the stored data in either a row or column configuration. A fixed-length instruction word can be decoded into two instructions, an operation instruction and an I/O instruction, and can be used to process the data. The I/O instruction can be disposed in an unused bit field of the operation instruction.

Abstract: The present invention is directed to a method of discriminating between textual content and graphical content. The method includes the steps of receiving a plurality of pixel values for a pixel line segment, calculating a plurality of spatial gradients based on the pixel values of adjacent pixels, determining a smoothness index by processing the plurality of spatial gradients, and identifying the pixel line segment as text or graphics by comparing the smoothness index to a threshold value.

Abstract: A stochastic screening mask is provided for rendering halftone images. A filter controlled force masking method is used to vary the mask patterns from regularity to randomness. Dot size (amplitude), dot number (frequency), and dot shape are modulated integrally at each gray level during the generation of the mask. The single pixel limit for blue noise masking is broken down by this approach. The halftone patterns generated with this method can have visually pleasing blue noise attribute. This new screening method would provide flexible solutions for different types of printing processes.

Abstract: A stochastic screening mask is provided for rendering halftone images. A filter controlled force masking method is used to vary the mask patterns from regularity to randomness. Dot size (amplitude), dot number (frequency), and dot shape are modulated integrally at each gray level during the generation of the mask. The single pixel limit for blue noise making is broken down by this approach. The halftone patterns generated with this method can have visually pleasing blue noise attribute. This new screening method would provide flexible solutions for different types of printing processes.

Abstract: A printing system and method performs lossy compression on a bitonal image when the printing system runs out of available memory for processing a page. The lossy compression is performed by dividing the image into subregions and by then determining a contone value for each subregion. Next, each contone value is converted to a bitonal value (i.e. either a "1" or a "0") to achieve lossy compression. In converting the contone values to bitonal values, the system determines the error associated with each conversion and diffuses the error to adjacent subregions using a random perturbation technique to determine the proportion of the error to diffuse to each subregion. When compressing color images, contone values are determined for each color plane. A relative color ordering of each contone value is also stored and used in converting the contone values to bitonal values in order to solve phasing and color alignment problems which would otherwise arise.