Abstract:
A method for using a scanner to scan pixel data from an image on a page and map the scanned pixel data in a computer's memory is disclosed comprising the steps of (a) placing a scanner having an elongated detector array of detector elements and two optical mice located in relation to the ends of the array on a transparent grid pattern covering an image on a page to be scanned, (b) determining the orientation of the scanner with respect to the grid pattern using motion detection information provided by the two mice, (c) positioning the scanner at a starting point from which to start detecting the position of the mice and commencing scanning of the page by moving the scanner across the surface of the grid pattern, and (d) storing pixel data scanned by each detector element at a memory address of a computer in communication with the scanner which corresponds to a coordinate position on the grid pattern determined to be nearest the actual position of a particular detector element.
Abstract:
A multiple camera automatic digitizer includes a plurality of spaced video cameras (19) mounted within a housing (3) under a transparent digitizing surface (51). Multiplexing circuitry (37) multiplexes video signals from a selected video camera into a video amplifier (51), the output of which is converted to a digital number representing the darkness of the point of the document most recently scanned by a presently selected one of the video cameras. In accordance with the invention, the automatic digitizer includes a microprocessor (111), system controls that allows selection of various ones of the video cameras to achieve digitizing of corresponding areas of a large document, and computes skew correction factors for each camera as the basis of digitizing of a plurality of permanent alignment marks (P) on a bottom surface of a cover (11) for the digitizing surface. The skew correction factors are used to skew correct the camera data produced by the corresponding camera. The skew corrected lateral offset of each camera is computed by the microprocessor on the basis of the digitized coordinates of the alignment points and is used to automatically compute skew corrected, offset corrected document coordinates for all of the video cameras.
Abstract:
An image joining method for a scanner that scans and transfers image data to a terminal is provided. During scanning and data transfer, when the image data stored in a register is full, an image processor stops the scanning of a linear photodetector. Meanwhile, a matrix photodetector fetches and stores the image of a code strip. Then, the photodetectors are moved backward a distance larger than that being required to be accelerated forward when the scanner resumes scanning. After the data in the register being transferred and cleared, the linear photodetector resumes scanning at a normal moving speed. The code strip image is further fetched and compared with the stored one to get a joining point of scanned image data. By the joining point, the new and prior image data portions are joined and the scanning proceeds, and joined image data will not be overlapped or broken.
Abstract:
There exists a problem that, for registration correction, carrying out less-than-one-pixel-basis correction and one-pixel-basis correction for an image after half toning using screen processing causes unevenness in density when it is hard for linearity of image density against a PWM to come out. Processing of interpolation is carried out so that the way how a halfdot is formed may always becomes the same after the interpolation processing. By applying a matrix in consideration of a dot direction which may be broken by the interpolation processing at the time of the screen processing, conversion in which a way of locating the half dot maintains relation of vertical inversion in a subscanning direction is carried out, and less-than-one-pixel-basis correction is carried out.
Abstract:
There exists a problem that, for registration correction, carrying out less-than-one-pixel-basis correction and one-pixel-basis correction for an image after half toning using screen processing causes unevenness in density when it is hard for linearity of image density against a PWM to come out. Processing of interpolation is carried out so that the way how a halfdot is formed may always becomes the same after the interpolation processing. By applying a matrix in consideration of a dot direction which may be broken by the interpolation processing at the time of the screen processing, conversion in which a way of locating the half dot maintains relation of vertical inversion in a subscanning direction is carried out, and less-than-one-pixel-basis correction is carried out.
Abstract:
A method of processing image data from a multi-chip array with a plurality of photosensitive chips aligned substantially in a transverse direction, including: generating, using a processor for at least one specially programmed computer, a Δy or Δx optical error value equal to a difference in process and transverse directions, respectively, between actual and apparent locations for a first photosensor, the apparent location due to optical error; and storing, in a memory element for the specially programmed computer, respective outputs from the photosensors in the array for first and second scan lines. The actual location is included in the first scan line. The processor retrieves, for use as at least part of useful image data for the first photosensor, the stored output of: the first photosensor for the second scan line for a Δy optical error, or a second photosensor for the first line for a Δx optical error . . . .
Abstract:
A printer not optically correcting bends and inclinations of scanning lines needs to execute such control as to electrically correct them. However, a conventional correction method has a problem that this correction causes an image defect such as an image streak or an uneven density in a specific area. A correction method of the present invention, when an input image has only one color, executes only correction by a second correction component which corrects distortions in a main scanning direction without executing correction by a first correction component which corrects bends and inclinations in a sub-scanning direction, and when an input image has two or more colors, executes both of correction by the first correction component and correction by the second correction component. This control can reduce the frequency of occurrence of an image defect.
Abstract:
An image reading apparatus includes a sensor section which includes a color line sensor which reads a color image and a monochrome line sensor which reads a monochrome image provided in parallel with and a specific distance away from the color line sensor, an input section which inputs a document image to the color line sensor and the monochrome line sensor, and a control section which, when reading an image from a document by use of the input section, starts the reading of an image by the color line sensor and the reading of an image by the monochrome line sensor with the same timing.
Abstract:
Image distortion is corrected in a color printer wherein print units for a plurality of colors are aligned to form a color image by superposing images formed by the print units. A quantity of image distortion is detected, and correction data of main scan address and subscan address are calculated and stored according to the detected image distortion for each main scan address. When input color image data are corrected, printing position is corrected in combination of address change with density interpolation. When the correction data exceeds the maximum correction range, the correction data are replaced by the maximum in the correction range to utilize the capability of the distortion correction of the printer. Further, in the image data correction, after the image data are converted to data having a smaller number of gradation levels, the data are delayed according to the serial arrangement of the print units. Then, the delayed data are converted again for printing.
Abstract:
An image processing device for processing an image read by an image scanner or similar image reading device is disclosed. The image processing device is capable of accurately detecting, without resorting to a range finding sensor or similar special sensing means, the bound portion of a spread book by recognizing the configuration of the bound portion out of a read image, and thereby accurately correcting the distortion of the portion of the image representative of the bound portion.