摘要:
Methods and apparatus for segmenting image data into windows and rendering a best-estimate outline delineating document windows may include adjusting a background white threshold differently than adjusting a white threshold normally used to classify pixels as white for video processing purposes. Such an independently arrived at threshold may be more suitable for the different task of differentiating between document background and document content.
摘要:
A method for dynamic range adjustment of image data of a captured image by determining a white point of an image. The method also involves determining a black point of the image, classifying pixels of the image, and determining an offset value for a pixel of the image based on the determined black point of the image and the determined classification of the pixel. Dynamic range adjustment of the image data is performed using the determined offset value for the pixels of the image and the determined white point of the image.
摘要:
A method and system for image processing, in conjunction with classification of images between natural pictures and synthetic graphics, using SGLD texture (e.g., variance, bias, skewness, and fitness), color discreteness (e.g., R_L, R_U, and R_V normalized histograms), or edge features (e.g., pixels per detected edge, horizontal edges, and vertical edges) is provided. In another embodiment, a picture/graphics classifier using combinations of SGLD texture, color discreteness, and edge features is provided. In still another embodiment, a “soft” image classifier using combinations of two (2) or more SGLD texture, color discreteness, and edge features is provided. The “soft” classifier uses image features to classify areas of an input image in picture, graphics, or fuzzy classes.
摘要:
A system and method classify a pixel of image data as one of a plurality of image types. A first image characteristic value for the pixel, a second image characteristic value for the pixel, a third image characteristic value for the pixel, and a fourth image characteristic for the pixel is determined. Some of these determinations may be resolution dependent. The values from these determination are utilized in assigning an image type classification to the pixel. Moreover, if at least one of the image characteristic values is greater than a predetermined threshold value the pixel is classified as a halftone peak value. The system includes a plurality of microclassifiers for determining a distinct image characteristic value of the pixel; a plurality of macroreduction circuits connected to the plurality of microclassifiers for performing further higher level operations upon the distinct image characteristic values of the pixel to produce reduced values; and a classification circuit to classify the pixel as an image type based on the reduced values from the macroreduction circuits. The system also includes a circuit to detect flat peaks without detecting multiple peaks and a rectangular blur filtering system.
摘要:
A method and system implements a dynamic error diffusion process. A grey level value representing a pixel is received. An image segmentation circuit determines an image characteristic of the pixel being processed. A threshold circuit thresholds the grey level value (pixel) and generates an error value as a result of the threshold process. A portion of the error value is diffused to adjacent pixels on a next scanline. The distribution of this error value is dynamic in that different sets of weighting coefficients are used based the image characteristic of the processed pixel. One set of coefficients are utilized in processing a pixel having a first image characteristic, and a second set of coefficients are utilized in processing a pixel having a second image characteristic.
摘要:
A system and method classify a pixel of image data as one of a plurality of image types. A first image characteristic value for the pixel, a second image characteristic value for the pixel, a third image characteristic value for the pixel, and a fourth image characteristic for the pixel is determined. Some of these determinations may be resolution dependent. The values from these determination are utilized in assigning an image type classification to the pixel. An auto-segmentation circuit enables classifying a pixel as a continuous tone and detecting a gradient in the window of pixels of which the pixel is a member. A filter enhances the image data of the pixel tone when a gradient indicating that image data within the window of pixels is changing at a rate greater than a predetermined threshold is detected or lowpasses the image data of the pixel tone when a gradient indicating that image data within the window of pixels is changing at a rate less than a predetermined threshold is detected.
摘要:
A system and method classify a pixel of image data as one of a plurality of image types. A first image characteristic value for the pixel, a second image characteristic value for the pixel ,a third image characteristic value for the pixel, and a fourth image characteristic for the pixel is determined. Some of these determinations may be resolution dependent. The values from these determination are utilized in assigning an image type classification to the pixel. Moreover, if at least one of the image characteristic values is greater than a predetermined threshold value the pixel is classified as a halftone peak value. The system includes a plurality of microclassifiers for determining a distinct image characteristic value of the pixel; a plurality of macroreduction circuits connected to the plurality of microclassifiers for performing further higher level operations upon the distinct image characteristic values of the pixel to produce reduced values; and a classification circuit to classify the pixel as an image type based on the reduced values from the macroreduction circuits. The system also includes a circuit to detect flat peaks without detecting multiple peaks and a rectangular blur filtering system.
摘要:
A method and apparatus to process an electronic image of a document that is previously divided into a plurality of classified tiled rectangular blocks. Each pixel within halftone blocks is analyzed to calculate a busyness measure value, preferably a sum of Laplacians although other range-type functions could be used. When the busyness value is less than a first predetermined threshold value, the video intensity level of the pixel is compared against a second predetermined threshold value and when greater than the second predetermined threshold value, the pixel is re-classified as being a Text document type to produce a crisp, clean background in the rendered image. When the video intensity level of the pixel is less than the second predetermined threshold value, the pixel is re-classified as being a Contone image type to minimize abrupt changes in pixel classification block boundaries, which would result in objectionable artifacts in the rendered image resulting from density or texture changes at the block boundaries. Selectively re-classifying individual pixels within blocks of tiled pixels that were previously classified as halftone pixels greatly reduces rectangular artifacts in the classification map. Such rectangular artifacts in the classification map are often translated by classification driven image processing into undesirable rectangular artifacts when the image is viewed or analyzed further. Certain classification transitions are thereby moved to coincide with contours and transitions in the image video, thereby hiding the classification transitions.
摘要:
The present invention provides a marker detection and thinning process for a photocopier which is executed during a single scan of a document and does not require a prescan of the document. The system realtime analyzes pixel image data, which is stored in a memory during the scanning process, to detect a marker line of a predetermined color and to thin the marker line. This analysis begins with image data stored in the memory that corresponds to a first pixel. The first pixel is associated with the scanline being presently scanned. It is determined if the image data of the first pixel represents a color that is equivalent to the predetermined color. Marker data is then stored in a buffer at a location corresponding to the first pixel if it is determined that the image data of the first pixel represents a color that is equivalent to a predetermined marker color. The data, stored in the buffer at a second pixel location, is analyzed. The second pixel is associated with a scanline that is prior in time to the scanline of the first pixel. It is determined if the data at the second pixel location is marker data. Also, the data stored in pixel locations adjacent to the second pixel location is analyzed to determine if this data is marker data. The marker data at the second pixel location is eliminated if it is determined, upon examining data of pixel locations adjacent to the second pixel location, that the elimination of this marker data would maintain the connectivity of the marker line. However, the data at the second pixel location is maintained if it is determined that, upon examining the data of pixel locations adjacent to the second pixel location, the elimination of this marker data would cause disconnectivity of the marker line.
摘要:
A printing system for rendering marks on a recording medium receives a multi-level grey scale pixel value representing a pixel having a first resolution. A screening circuit generates a screened multi-level grey scale pixel value equal to (G.sub.L -V.sub.i)+(S.sub.i -Th)*Dmp.sub.vi *Mod.sub.Eff wherein G.sub.L is the maximum grey level value of the pixel, V.sub.i is equal to the multi-level grey scale pixel value of the first resolution, S.sub.i is equal to a screen value corresponding to a position of the pixel, Th is the threshold value, Dmp.sub.vi is a video dependent dampening factor, and Mod.sub.Eff is a modulation multiplication factor. An interpolator converts the screened multi-level grey scale pixel value to a second resolution, the second resolution being higher than the first resolution, and a binarization circuit binarizes the converted multi-level grey scale pixel value so as to output a binary signal and an error value, the error value having a resolution equal to the first resolution. The error value is diffused to multi-level grey scale pixel values corresponding to pixels adjacent to the pixel having the first resolution, and the binary signal is converted into a mark on the recording medium.