JP2008294831A - Image distortion correction device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that not only sufficient correction can not be performed but image quality is deteriorated when distortion of a scanned image is corrected in the conventional distortion correction processing method. <P>SOLUTION: The image distortion correction device which corrects distortion of the scanned image is provided with: an image decision means for deciding whether the scanned image is a character image or a photograph image by every block with predetermined size; a font information acquisition part which acquires font information to a block decided as the character image in an image inversion means; a skew angle detection means for comparing the font information with the scanned image to detect a skew angle for every block; a distortion correction performance means for performing distortion correction for every block according to the skew angle detected by the skew angle detection means; a vector image conversion part which converts the font information into a vector image; and a means for developing the vector image with a block with the predetermined size into raster image data. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image distortion correction apparatus.

  Many originals read using a flatbed scanner are sheet-like originals, and an openable / closable pressure plate is provided on the contact glass. After placing the original on the contact glass, the pressure plate is closed and the original is scanned. Yes. However, the original is not limited to a sheet, and book originals (books, booklets, etc.) may be handled as originals. In such cases, the book original is placed on the contact glass and the original is scanned. Will do.

  However, when a book document is used as a document, the page binding portion of the book document is lifted from the contact glass. In this way, if the page binding part of the book document is lifted from the contact glass, the page binding part is moved away from the focal plane, so the scanned image of the lifted part has image distortion, shadow, character blur, etc. Image degradation occurs. The page binding portion of the deteriorated image is difficult to read, and the recognition rate when performing character recognition processing by OCR is significantly reduced. In particular, the ratio is high in thick bookbinding, and the book document itself may be damaged when the pressure is applied so that the page binding portion of the book document is not separated from the focal plane.

  In order to solve such a problem, a method of correcting image distortion using a method of estimating the three-dimensional shape of an object from image density information has been proposed. As a method for estimating the three-dimensional shape of an object from such image density information, a method called Shape from Shading described in Non-Patent Document 1 is a typical example.

  However, according to the method called “Shape from Shading” described above, the amount of calculation is large, and the calculation time of the distortion correction processing is long.

  Further, according to the method described in Japanese Patent Application Laid-Open No. 5-110002, a special shape measuring device for measuring the shape of a book by a triangulation method is required, which is not appropriate.

  Therefore, in recent years, an image correction apparatus that estimates the three-dimensional shape of the book surface using the shape of the page outline of the read scan image has been proposed in order to effectively correct the distortion with a small amount of calculation (for example, a special feature). (See Kaihei 11-41455).

FIG. 16 shows a diagram illustrating an example of a conventional MFP system.
Japanese Patent Laid-Open No. 5-161002 JP 11-41455 A T. Wada, H. Uchida and T. Matsuyama, "Shape from Shading with Interreflections under a Proximal Light Source: Distortion-Free Copying of an Unfolded Book", International Journal Computer Vision 24 (2), 125-135 (1997)

  By the way, when setting a book document on the contact glass, it is necessary to set the book document image face down, so that the operator can only see the back side of the document. Therefore, as shown in FIG. 17, the scan image includes a scan image a in which the page binding portion of the book document is inclined with respect to the main scanning direction, which is so-called skew, and a book document as shown in FIG. Since the method of lifting the page binding part (distance from the contact glass) differs between the top and bottom of the page, a scan image b having a shape like a “C” may occur.

  However, since the conventional distortion correction processing method does not take these situations into consideration, there is a problem that not only the correction cannot be performed sufficiently but the image quality deteriorates if the distortion of the scanned image is corrected. there were.

In order to solve the above-described problems, the first aspect of the present invention reads a book document that contacts the top or bottom of the contact glass with the page binding portion parallel to the main scanning direction of image reading by the image reading means. In the image distortion correction device for correcting the distortion of the scanned image,
Image determination means for determining whether the scanned image read by the image reading means is a character image or a photographic image for each block of a predetermined size;
A font information acquisition unit that acquires font information for a block that is determined to be a character image by the image inversion means;
A skew angle detection unit that compares the scanned image with the font information acquired by the font information acquisition unit, and detects a skew angle for each block;
Distortion correction execution means for correcting distortion for each block according to the skew angle detected by the skew angle detection means;
A vector image conversion unit that converts the font information acquired by the font information acquisition unit into a vector image;
Means for expanding the vector image of the block of the predetermined size into raster image data.

  As described above, according to the present invention, the distortion of a scanned image obtained by reading a book document contacting the top or bottom of the contact glass with the page binding portion parallel to the main scanning direction of image reading is read by the image reading unit. In the image distortion correction apparatus for correcting image, the image determination means for determining whether the scanned image read by the image reading means is a character image or a photographic image for each block of a predetermined size, and the image inversion means The font information acquisition unit that acquires font information for a block that is determined to be a character image in step 1 and the font information acquired by the font information acquisition unit are compared with the scanned image, and a skew angle is detected for each block. According to the skew angle detected by the skew angle detecting means and the skew angle detected by the skew angle detecting means. A distortion correction executing means for performing distortion correction for each block; a vector image converting unit for converting the font information acquired by the font information acquiring unit into a vector image; and a vector image of the block of the predetermined size is rasterized. And a means for developing the image data, it is possible to obtain a skew angle for each block obtained by dividing the scan image into predetermined sizes, and to perform an optimal distortion correction process for each block, thereby achieving more accurate accuracy. High distortion correction processing can be performed.

<Embodiment 1>
FIG. 1 is a block diagram showing the configuration of the image distortion correction apparatus according to the first embodiment of the present invention.

[Outline of MFP equipment]
Hereinafter, the apparatus of the present invention and its operation will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram showing a first embodiment of an MFP controller according to the present invention.

  The MFP controller 1 is connected to the CPU 3, the memory controller (MC) 4, the general-purpose bus 6, the tile tile page vector conversion unit 13, the raster → tile vector conversion unit 14, and the tile vector development unit (RIP) 18 through a system bus bridge (SBB) 2. Has been. The RIP 18 is composed of a plurality of small development parts (μRIPa to d).

  A system memory (Memory) 5 is connected to the tip of the MC 4 and is used as a medium for temporarily storing image data.

  The general-purpose bus 6 includes a hard disk controller (HDDCont) 7 that controls the HDD 8 for storing image data, an operation unit controller (LCDC) 9 that controls the operation unit (LCD) 10, and a network 12 to which the MFP is connected. The LAN I / F 11 serving as an interface for transferring image data between external devices is connected.

  An image processing unit 15 is connected to the tip of the raster → tile vector conversion unit 14, and a scanner 16 and a printer 17 are connected to the image processing unit 15.

  In addition, a RIP 18 is connected to the SBB2, and a local memory (LocalMemory) 19 for storing data output from the RIP 18 is connected to the SBB2.

  The image data handled by the controller 1 is interfaced with page vectors (PDL, PDF, SVG, etc.) for input / output with an external device, and raster data for input / output with a scanner or printer. In the controller 1, unlike a conventional MFP, the scanned image is converted into a tile vector by a tile vector / raster raster → tile vector conversion unit 14. The DL data is stored in the local memory 18 connected to the RIP 18. Accordingly, only two types of images of page vector and tile vector data are stored on the system memory 103. That is, since it is not necessary to store raster data and DL data having a large image size in the system memory 5, it is possible to reduce an image data area that must be secured on the system memory.

  In addition, since the DL data output from the RIP 18 is stored as DL data divided into tile units, it can be stored with a very small memory capacity compared to conventional DL data in page units. Therefore, the local memory 19 can be mounted on-chip, and the memory latency can be reduced. As a result, the tile data development speed can be increased. In addition, since only the tiled data needs to be stored on the HDD 8 as image data, the bottleneck of the access speed to the HDD is alleviated and the data processing can be speeded up. At the same time, the cost of RIP can be reduced by tiling.

  When a higher processing capability is required, the processing capability can be varied by mounting a plurality of μRIPs provided in the RIP in parallel. That is, since the processing capacity of the controller can be simply adjusted, a system that can easily ensure scalability can be constructed.

  Hereinafter, each image processing flow when the MFP according to the present invention is operated will be described.

[copy]
FIG. 2 shows a data flow in the MFP when the copy operation is performed. The arrows drawn in FIG. 2 represent the flow of data. An arrow drawn with a solid line means a raster image, an arrow drawn with a broken line means a tiled vector image, and an arrow drawn with an alternate long and short dash line means that a vector image describing the entire page flows. The page vector and tile vector image will be described in detail in a tile-page vector conversion unit described later.

(S21):
When the user gives an instruction to start copying from the operation unit (LCD) 10, the scanner 14 starts a document image reading operation.

  The image (R, G, B) input from the scanner 14 to the image processing unit 13 is subjected to the following processing after being frequency-converted to the clock synchronization of the image processing block.

  Correction of scanner characteristics such as CCD sensor line pitch and chromatic aberration.

  Image quality correction of input image data such as color space correction and sharpness.

  Image processing such as erasing input image data and erasing book frames.

(S22):
When the image processing is completed, the image data output from the image processing unit 13 is input to the raster-to-tile vector conversion unit 12 to perform tile vector conversion processing. That is, the image data is divided into blocks of a predetermined size, and a vectorization process is performed on the raster image in each block to generate a vector image in units of blocks.

The generated vector image is subjected to bus arbitration by the SBB 2, acquires the bus right to the system memory 5, and stores the tile vector in the system memory 5 through the MC 4. (Note that when a data path is connected via SBB2, the procedure for acquiring the bus right is basically received after bus arbitration, but this is omitted in the subsequent flow description.)
(S23):
The tile vector image stored in the system memory 5 is stored in the HDD 8 via the SBB 2 via the HDDCont 7 and the MC 4. By storing image data in the HDD 8, sorting can be performed when copying a plurality of originals, and the pages can be output in a different order or stored as a saved document in the MFP.

(S24):
The tile vector image stored in the HDD 8 is read out by the HDDCont 7 in accordance with the printer ready timing sent from the CPU (not shown) in the printer 15 and temporarily stored in the system memory 5 via the SBB 2 and MC 4. Is done.

  If the read image data is directly read from the HDD 8 to the printer, it cannot be guaranteed that the access speed of the HDD 8 becomes regular, or that it is output in synchronization with the printer due to the degree of congestion of the general-purpose bus 6. Therefore, spooling page data in the system memory 5 before data transfer in synchronization with the printer guarantees real-time throughput.

(S25):
The tile vector image stored in the system memory 5 is read by the MC 4 in accordance with an activation signal sent from the printer 15 to the controller 1 and transferred to the RIP 18 via the SBB 2.

  First, the RIP 18 analyzes a tile vector and generates (interprets) a drawing object (tile DL data) in units of tiles. The generated tile DL data is temporarily stored in the local memory 18.

  The RIP 18 reads the tile DL data from the local memory, develops it into a raster image in units of tiles, and outputs it.

  In this embodiment, as described above, the RIP 18 includes four μRIPa to μRIPd. The controller 1 operates μRIPa to μRIPd in parallel, thereby expanding the tile vector at high speed. The system performance is dominated by the vector expansion time, and the performance can be improved by increasing the μRIP. Therefore, it is possible to easily construct a scalable system by using the configuration of the present invention.

(S26):
The image data rasterized in tile units by the RIP 18 is transferred to the image processing unit, and the following processing is executed.

  Conversion processing from a tile unit raster image to a page unit raster image.

  Output color and density correction processing that matches the printer characteristics.

  Halftone processing that quantizes image data and performs tone conversion of the output image.

  Frequency conversion processing to output images in synchronization with the printer I / F clock.

  The raster image data subjected to the image processing by the image processing unit 15 is transferred to the printer 15, printed on a recording medium, and output.

[Print]
FIG. 3 shows a data flow in the MFP when the printing operation is performed.

(S31):
The LAN I / F 11 connected to the general-purpose bus 6 receives page vector format image data from an external device connected to the network 12. Then, the data is transferred to the system memory 5 via the MC 4 connected to the end of the SBB 2.

(S32):
The page vector image stored in the system memory 5 is read out from the tile / page vector conversion unit 13 and performs tile vector conversion processing. That is, an object existing in a page vector is divided into objects that can be contained in a block (tile) having a predetermined size, and a vector image in units of tiles is generated.

(S33):
The generated vector image is stored again in the system memory 5 via the SBB 2.

(S33):
The tile vector image stored in the system memory 5 is stored in the HDD 8 via the SBB 2 via the HDDCont 7 and the MC 4. By storing image data in the HDD 8, sorting can be performed when copying a plurality of originals, and the pages can be output in a different order or stored as a saved document in the MFP.

(S34):
The tile vector image stored in the HDD 8 is read out by the HDDCont 7 in accordance with the printer ready timing sent from the CPU (not shown) in the printer 15 and temporarily stored in the system memory 5 via the SBB 2 and MC 4. Is done.

  When the read image data is read directly from the HDD 8 to the printer, it cannot be guaranteed that the access speed of the HDD 8 becomes regular, or that the output is synchronized with the printer due to the degree of congestion of the general-purpose bus 6. Therefore, before transferring data in synchronization with the printer, the tile vector data for one page is spooled in the system memory 5 to guarantee real-time throughput.

(S35):
The tile vector image stored in the system memory 5 is read by the MC 4 in accordance with an activation signal sent from the printer 15 to the controller 1 and transferred to the RIP 18 via the SBB 2.

(S36):
First, the RIP 18 analyzes a tile vector and generates (interprets) a drawing object (tile DL data) in units of tiles. The generated tile DL data is temporarily stored in the local memory 18.

  The RIP 18 reads the tile DL data from the local memory, develops it into a raster image in units of tiles, and outputs it. In the present embodiment, the RIP 18 includes four small development portions μRIPa to μRIPd.

  The tile vectors are developed at high speed by operating the μRIPa to μRIPd in parallel. The system performance is dominated by the vector expansion time, and the performance can be improved by increasing the μRIP. Therefore, it is possible to easily construct a scalable system by using the configuration of the present invention.

(S37):
The image data rasterized in tile units by the RIP 18 is transferred to the image processing unit, and the following processing is executed.

  Conversion processing from a tile unit raster image to a page unit raster image.

  Output color and density correction processing that matches the printer characteristics.

  Halftone processing that quantizes image data and performs tone conversion of the output image.

  Frequency conversion processing to output images in synchronization with the printer I / F clock.

  The raster image data subjected to the image processing by the image processing unit 15 is transferred to the printer 15, printed on a recording medium, and output.

[Send]
FIG. 4 shows the data flow in the MFP when it is transmitted over the network. The flow until image data is stored in the HDD 8 is omitted because it is the same as [Copy] for a scanned image and [Print] for an image input from an external device on the network.

(S41):
The tile vector image stored in the HDD 8 is read from the HDDCont 7 connected to the general-purpose bus 6 via the SBB 2 and temporarily stored in the system memory 5.

(S42):
The tile vector image stored in the system memory 5 is read from the tile / page vector conversion unit 11 and performs tile vector conversion processing. That is, the objects divided into blocks are combined to generate a page vector image describing the object in the entire page.

(S43):
The generated page vector image is stored again in the system memory 5 via the SBB 2.

(S44):
The page vector image stored in the system memory 5 is read from the LAN I / F 11 connected to the general-purpose bus 6 and transferred to an external device connected to the network 12.

  As in the present invention, when transmitting to an external device, the tile vector image is returned to the page vector image and the number of objects is reduced, so that the amount of transmission data can be reduced. In addition, it can be easily converted to general-purpose formats such as PDF and SVG.

[Raster-Tile Vector Converter]
Details of the raster-to-tile vector conversion unit 14 in the controller 1 will be described. FIG. 5 shows an internal processing flow of the raster tile conversion unit 14.

  The raster → tile vector conversion unit 14 is executed by the following steps.

(S51) BS (block selection):
The raster image data input from the image processing unit 15 is divided into a character / line drawing area including characters or line drawings, a halftone photo area, an irregular image area, and the like. Further, with respect to the character / line drawing area, a character area mainly including characters and a line drawing area mainly including tables and figures are separated, and the line drawing area is separated into a table area and a graphic area. In the present embodiment, connected pixels are detected and separated into regions for each attribute using the shape, size, pixel density, etc. of the circumscribed rectangular region of the connected pixels.

  The character area is segmented into rectangular blocks (character area rectangular blocks) using a cluster of each character paragraph as a block. In the line drawing area, each object (table area rectangular block, line drawing area rectangular block) such as a table or a figure is segmented into rectangular blocks.

  A photo area expressed in halftone is segmented into rectangular blocks for each object such as an image area rectangular block and a background area rectangular block.

  Each separated region is further divided into regions (tiles) of a predetermined size, and vectorized in the next vectorization step in units of tiles.

(S52) Vectorization step:
Image data of each attribute area is converted into vector data by vectorization processing.

  In the vectorization process, two types of vectorization processes are executed: a drawing or line other than the character area, a vector area process of the table area, and a vector area type of the character area. Further, vectorization is not performed on the photograph area as image data as it is.

  Thus, first, vectorization processing of a graphic, line, or table area other than a character block will be described with reference to FIG.

  FIG. 6 is a flowchart showing details of vectorization processing of a graphic, line, or table area other than the character block according to the first embodiment of the present invention.

  In the vectorization process of a graphic or line other than a character block, or a table area, first, in step S1001, based on image data corresponding to the area to be processed, the area is expressed as a connection of line segments (line information). Hereinafter, an outline vector conversion process for converting into an outline vector) is executed.

  Next, in step S1002, a graphic recognition process is executed from the obtained outline data.

"Figure recognition processing"
In the figure recognition process, the outline data obtained in step S1001 is extracted from the outline by recognizing a figure shape such as a circle, an ellipse, a square, or a triangle, or a figure shape such as a ruled line or a curve formed by the outline. Recognize and extract.

  Next, details of the vectorization processing of the character area will be described with reference to FIG.

  FIG. 7 is a flowchart showing details of the vectorization processing of the character area according to the present invention.

  First, in step S1011, character recognition processing for a character block is executed. In step S1012, the character block font recognition process is executed. In step S1013, a character vector conversion process that outlines a character block is executed. In step S1014, font conversion processing is executed on the character block.

  Here, the details of the character recognition processing in step S1011 will be described.

"Character recognition processing"
In the character recognition process, character recognition is performed on a character image cut out in character units from a character block using a pattern matching technique, and a corresponding character code is acquired. In particular, this character recognition process compares an observed feature vector obtained by converting a feature obtained from a character image into a numerical sequence of tens of dimensions and a dictionary feature vector obtained for each character type in advance, and has the closest distance. The character type is the recognition result.

  There are various known methods for extracting a feature vector. For example, there is a method characterized by dividing a character into meshes and using a mesh number-dimensional vector obtained by counting character lines in each mesh block as line elements according to directions.

  When character recognition processing is performed on a character block, first, horizontal / vertical writing is determined for the corresponding character block, a character string is cut out in each corresponding direction, and then a character is cut out from the character string. Get a character image.

  The horizontal / vertical writing is determined by taking a horizontal / vertical projection of the pixel value in the corresponding character block. If the horizontal projection has a large variance, the horizontal writing is determined, and if the vertical projection has a large variance, the vertical writing is determined. If the block is a horizontally written character block, the character string and character are decomposed by cutting out the line using the horizontal projection and cutting out the character from the vertical projection of the cut line. . On the other hand, for vertically written character blocks, horizontal and vertical may be reversed.

  Note that the character recognition rate (pattern matching matching rate) can be obtained by this character recognition processing.

  Next, details of the font recognition processing in step S1012 will be described.

"Font recognition process"
Font recognition processing prepares a plurality of dictionary feature vectors for the number of character types used for character recognition processing for character shape types, that is, font types, and outputs font types together with character codes at the time of matching. By recognizing the character font.

  Next, details of the character vector conversion processing in step S1013 will be described with reference to FIG.

"Character vector conversion process"
FIG. 8 is a flowchart showing details of the character vector conversion processing of the present invention.

  The processing unit in FIG. 8 is a character image for one character divided from the character area by the character recognition process, and the processing in FIG. 8 is executed for all the character images in the character area.

  First, in step S1020, it is determined whether the character recognition rate obtained by the character recognition process exceeds a predetermined recognition rate. If it is determined that the character recognition rate exceeds the predetermined recognition rate (YES in step S1020), the process proceeds to step S1021, and character outline data is prepared using font outline data prepared in advance corresponding to each character image. Get the font information of the part and convert it to vector data.

  On the other hand, when it is determined that the character recognition rate is lower than the predetermined recognition rate (NO in step S1020), the process proceeds to step S1022, and a distortion amount (skew angle) is obtained. Specifically, the image of the character part is rotated by several steps of angle, character recognition processing is performed on the rotated image, and the rotation angle with the highest character recognition rate is set as the distortion amount (skew angle).

  Thereafter, the font information of the character part is acquired in step S1023, and the process proceeds to step S1024.

  In step S1024, it is determined whether character vector conversion has been completed for all characters in the character area. If it is determined that the character vector conversion has been completed for all characters, the process proceeds to step S1025. If it is determined that the character vector conversion has not been completed, the process proceeds to step S1020.

  In step S1025, the amount of distortion of the character area is calculated. Specifically, the distortion amount of each character obtained in step S1022 is summed, and the distortion amount (skew angle) of the character area is obtained by dividing the total distortion amount by the number of characters advanced to step S1022. If there is no character proceeding to S1022, the distortion amount of the character area is set to zero.

  Next, details of the font conversion processing in steps S1021 and S1023 will be described.

"Font processing"
In the font processing, each character image is replaced with font data using outline vector data or image data obtained by the character vector conversion processing. By replacing the character image with font data, the character block can be expressed as a sentence composed of each character and can be edited.

  Specifically, the outline vector data obtained in steps S1021 and S1023 can be replaced with a predetermined font using the character code, font information, and character size as they are.

(S53) Tile vector generation:
The vector type for determining whether the data is vector-converted into command definition information such as a to f format code information, graphic information and function information in S52, and the page type or tile vector in the controller 1 The header information for discriminating the coordinate position is added. In this way, tile vector data packed in units of tiles is output to SBB2.

(S55) Distortion correction processing:
Distortion correction processing is performed on the tile data for one page generated in S53. This distortion correction processing is executed for each tile.

  FIG. 9 is a flowchart showing details of the distortion correction processing in step S55. The distortion correction process will be described with reference to FIG.

  First, in step S1030, it is determined whether the area (tile) is a photographic area. If it is determined that the area is a photograph area, the process proceeds to step S1031, and if it is determined that the area is not a photograph area, the process proceeds to step 1033.

  Next, in step S1031, the amount of distortion (skew angle) of the photographic area is calculated. Specifically, the distortion amount of the character area closest to the photographic area is set as the distortion amount of the photographic area. When there are a plurality of character areas closest to the photo area, the average distortion amount of the character areas is set as the distortion amount of the photo area.

  Subsequently, in step S1032, distortion correction processing is performed. Specifically, the image in the photographic area is rotated by the skew angle obtained in step S1031.

  In step S1033, it is determined whether or not the distortion correction processing has been completed for all tiles. If it is determined that the distortion correction process has not been completed for all tiles, the process proceeds to step S1030. If it is determined that the distortion correction process has been completed for all tiles, the distortion correction process of S55 is terminated. To do.

[Tile tile page vector conversion part]
Details of the tile tile page vector conversion unit 13 in the controller 1 will be described.

  FIG. 12 shows a document created by an application of an external device on the network. For convenience, the short direction of the document is defined as the 'X' direction, and the long direction is defined as the 'Y' direction.

  FIG. 13 shows a description example of a page vector (PDL command) for instructing printer output of the document of FIG.

  In FIG. 13, reference numeral 901 denotes a document setting command relating to setting of the entire document, 902 denotes a character drawing command, and 903 denotes a graphic drawing command.

  Details of the drawing commands 901 to 903 will be described.

  C1 to C5 are commands related to the entire document. Therefore, these commands C1 to C5 have only one place for one document. These commands related to the entire document include, for example, character set commands (font specification commands), scalable font commands (commands that specify whether or not to use scalable fonts), and hard reset commands (reset the previous printer environment) Command), etc. C1 is a document setting start command. C2 is a command representing the output paper size of the document, and in this case, A4 is set. The next C3 indicates the direction of the document. There are portrait and landscape, but in this case, portrait (PORT) is shown. C4 is a command representing a document type, which indicates whether the document is a page vector or a tile vector. In this case, it is a page (PAGE). C5 is a document setting end command.

  C6 to C21 are commands for outputting the document 801. C6 is for indicating the start of a page. C7 is a command for selecting the font type of the character, and in this case, the font set numbered "1" is selected. C8 sets the font size, and the size of “10 points” is selected. C9 is a command for setting the color of the character, and indicates the brightness of each color component of R (red), G (green), and B (blue) in order. It is assumed that this luminance is quantized in 256 steps from 0 to 255. C10 indicates the coordinates of the start position for drawing the character. The coordinate position is specified with the upper left corner of the page as the origin. In this case, the character drawing is set to start from the position {10, 5}. C11 indicates a character string (XXXX ...) to be actually drawn.

  C12 indicates the fill color of the surface at the time of drawing the figure. The color specification is the same as the character color. C13 designates the line color of the graphic drawing. C14 indicates the coordinates of the position where the figure is drawn. C15 is a command for designating a radius for drawing an arc, and in this case represents a "10" coordinate unit. C16 is for drawing a closed arc. Two parameters in the command indicate a drawing start angle and an end angle when drawing an arc. The vertical information is assumed to be 0 degree, and in this case, an arc of 0 degree to 90 degrees is drawn. C17 to C21 specify the surface and line color, the position, etc., as in the commands from C12 to C16. C22 specifies the end of the command.

  FIG. 11 shows an example in which the document of FIG. 12 is described by a tile vector in block units as shown in FIG. The two arrows in FIG. 14 represent the short direction “X” and the long direction “Y” of the document. In the drawing, the number sequence arranged in the X direction represents the tile ID in the X direction, and the number sequence arranged in the Y direction represents the tile ID in the Y direction. A, B, C, and D represent tile vectors at the positions of tile IDs (0, 0), (0, 1), (2, 4), and (1, 5), respectively.

  In FIG. 15, reference numeral 1101 denotes a document setting command relating to setting of the entire document, 1102 denotes the entire drawing command, and 1103 to 1106 denote drawing commands for tiles A, B, C, and D. Reference numerals 1107 and 1108 denote a character drawing command and a graphic drawing command for the tile D, respectively.

  Details of the rendering commands 1101 to 1108 will be described below.

  C1 to C5 are commands related to the entire document. Therefore, these commands C1 to C5 have only one place for one document. These commands related to the entire document include, for example, character set commands (font specification commands), scalable font commands (commands that specify whether or not to use scalable fonts), and hard reset commands (reset the previous printer environment) Command), etc. C1 is a document setting start command. C2 is a command representing the output paper size of the document, and in this case, A4 is set. The next C3 indicates the direction of the document. There are portrait and landscape, but in this case, portrait (PORT) is shown. C4 is a command representing a document type, which indicates whether the document is a page vector or a tile vector. In this case, it is a tile (TILE). C5 is a document setting end command.

  C6 to C500 are commands for outputting the document 1001. C6 is for indicating the start of a page. C7 indicates the start of the drawing command for tile A in FIG. Two parameters indicate the ID of the tile in the page. C8 represents the end of the tile A drawing command. When there is no object as in tile A, only the start and end of the tile are described.

  C9 indicates the start of the drawing command for tile B in FIG. C10 is a command for selecting the font type of the character. In this case, the font set numbered “1” is selected. C11 sets the font size, and the size of “10 points” is selected. C12 is a command for setting the character color, and indicates the luminance of each color component of R (red), G (green), and B (blue) in order. It is assumed that this luminance is quantized in 256 steps from 0 to 255. C13 indicates the coordinates of the start position for drawing the character. C14 indicates a character string (XXXX) to be actually drawn. The coordinate position is specified with the upper left corner of the tile as the origin. In this case, the character drawing is set to start from the position {0, 5}. C15 indicates the end of the drawing command for tile B.

  C100 indicates the start of the drawing command for tile C in FIG. C101 indicates the fill color of the surface when drawing a figure. The color specification is the same as the character color. C102 designates the line color of the figure drawing. C103 indicates the coordinates of the position where the figure is drawn. C104 is a command for designating a radius for drawing an arc, and in this case represents "10" coordinate unit. C105 is for drawing a closed arc. Two parameters in the command indicate a drawing start angle and an end angle when drawing an arc. The vertical information is assumed to be 0 degree, and in this case, an arc of 0 degree to 90 degrees is drawn. C106 indicates the end of the drawing command for tile C.

  C120 to C131, like the commands C9 to C15, specify the font type, color, size, etc. for character drawing, and specify the drawing surface, line color, position, etc., as in C100 to C106. ing. C500 designates the end of the page, that is, the end of the command.

  The tile tile page vector conversion unit 13 converts the page vector and the tile vector as described above. FIG. 10 shows an internal processing flow of tile-> page vector conversion.

  The tile-page vector conversion is executed in the following steps.

(S601):
First, a command sequence corresponding to the header portion is read from the vector image stored in the system memory 5, and the command portion relating to the entire document is analyzed. This corresponds to C1 to C5 in FIG. 13 or FIG.

(S602):
As a result of the analysis, if the document type is a page vector (PAGE), the process proceeds to step S603 and subsequent steps, and page vector → tile vector conversion is executed. If the document type is (TILE), the process proceeds to step S610 and subsequent steps, and tile vector → page vector conversion is executed.

(S603):
Read a command sequence that describes an object.

(S604):
The command sequence read in S603 is analyzed, and it is determined whether or not the size of the described object exceeds the tile size to be divided. If the object is not divided, step S605 is skipped and the process proceeds to step S606. If the object is divided, the process proceeds to step S605.

(S605):
In this step, the input object is divided.

  For example, in FIG. 13, drawing commands for all character strings including “XXXX” are described at 902, but when tiled as shown in FIG. 14, only “XXXX” can be contained in the tile B. Therefore, in the tile vector, the character string is divided in the middle, and the divided subsequent character string is described in the next tile as another character string. If the description does not fit in the next tile, it is similarly divided into character strings included in the tile, and the process is repeated until all the divided character strings fit within the tile size. To determine where to cut the character string, the number of characters that fit in the tile is calculated from the type and size of the font, and that many characters are extracted. In 902, the number of characters is four, and the character string of C10 in FIG. 13 is converted into a description as C13 in FIG.

  Further, although the drawing of the figure is described in 903, the figure described in C17 to C21 in 903 has a tile D because the three-quarter circle in FIG. 12 does not fit in one tile in FIG. Divided into multiple tiles. To divide a figure, calculate the part that touches the boundary area of the tile from the drawing position of the figure, the shape and size of the figure, and create a new closed area consisting of the boundary and the partial area of the figure that is contained in the tile. Rewrite as. In FIG. 15, the three-quarter circle described in 903 in FIG. 13 is converted into a description of a quarter circle such as C126 to C130 in the lower left partial region in FIG. The remaining area is also converted into a quarter-circle description of the same shape.

(S606):
In order to change the drawing description in the tile vector for the command description of the input object, the coordinate position is converted. The page vector describes the position from the upper left of the page, whereas the tile vector rewrites the position from the upper left of the tile. By describing the drawing position with the coordinates in the tile, the data length required for coordinate calculation can be reduced.

(S607):
When the command description conversion for one object is completed, it is determined whether or not the command description conversion for all objects in the page is completed. If not, the process returns to S603 and the next command is processed. The processing of S603 to S607 is repeated. When the command processing for one object is completed, the process proceeds to S608.

(S608):
When the description conversion of all the drawing commands is completed, the tile areas divided as shown in FIG. 14 are written to the system memory 5 as tile vectors in order from the upper left of the page. The tile vector is described in a format in which commands indicating the start and end of tiles are added to the commands described in S605 and S606 described above.

  First, at the time of writing the first command of the page, a tile vector in which no object is present in the system memory 5 is generated. A tile vector having no object, for example, tile A in FIG. 15, is described as C7 to C8 with only the start and end commands as shown in 1103. Next, a description of the object is added to the tile of the coordinate where the command processed in S603 to S607 exists. In the case of the tile B, the description is as C9 to C15 of 1104 in FIG. When a plurality of objects exist in the same tile as in the tile D, the object description 1107 and the object description 1108 are listed as 1106.

(S609):
When the writing of one object to the tile vector is completed, it is determined whether or not the description of all the objects on the page has been completed. If it is determined that the process has not ended, the process returns to S603. When all the processes in the page are finished, the page → vector tile conversion is finished.

  A process when it is determined in S602 that the document type is a tile (TILE) will be described.

(S610):
Read a command sequence that describes an object.

(S611):
The command sequence read in S610 is analyzed, and it is determined whether or not the described object can be combined with tiles read before that time. If the objects are not to be combined, S612 is skipped and the process proceeds to S613. If the objects are to be divided, the process proceeds to S612.

(S612):
Whether or not the objects can be combined is determined based on the coordinate position of the read command, the type of figure, and the like. In the case of a character string, the determination is made based on the font size and font type. Basically, the connection is performed in a manner that the flow of S605 is reversed.

(S613):
In order to change the drawing description in the tile vector for the command description of the input object, the coordinate position is converted. In the tile vector, the position from the upper left of the tile is described, whereas in the page vector, the position is described from the upper left of the page.

(S614):
When the command description conversion for one object is completed, it is determined whether or not the command description conversion for all the objects in the tile is completed. If not, the process returns to S610 and the next command is processed. The processes of S610 to S613 are repeated. When the command processing for one object is completed, the process proceeds to S615.

(S615):
When the drawing command description conversion is completed, the page vector is written to the system memory 5. The tile vector is described in a format in which commands indicating the start and end of tiles are deleted from the commands described in S605 and S606.

  First, when a command written in the first tile in the page is written, a page vector in which no object is present in the system memory 5 is generated. In FIG. 9, the page is described only by C1 to C6 and C22. Next, a description of the object processed in S610 to S613 is added. In FIG. 13, the description portion C7 to C11 in 902 corresponds to that. The object in this case represents the character string {XXXX... YY..., Which is the description of the object in which the character strings of the tiles described in 1104, 1107, etc. in FIG. has been edited.

(S616):
When the writing of one command to the page vector is completed, it is determined whether or not the description of all the objects of the tile has been completed. If it is determined that the process has not ended, the process returns to S610. When all the processes in the tile are completed, the process proceeds to S617.

(S609):
When the writing of one tile vector is finished, it is determined whether or not all the tile descriptions on the page are finished. If it is determined that the process has not ended, the process returns to S610. When the processing of all tiles in the page is completed, the tile → page vector conversion ends.

[Tile vector development (RIP)]
Details of the RIP 18 in the controller 1 will be described.

  First, before starting image data processing such as copying, printing, and transmission, the local memory 19 is initialized and the resolution of a drawing object to be created is set. In this embodiment, the generation resolution is 600 dpi, and the print command specified in the unit system such as the point size or mm is converted into the number of dots using this value.

  Expansion of tile vectors is performed in the following steps.

(S71):
Tile vectors input from the system memory 5 to the RIP 18 through the SBB 2 by a certain size are temporarily stored in the tile vector area of the local memory 19.

(S72):
When the tile vector is taken into the local memory 19, μRIPa to d determine whether or not the tile development process has been completed. If μRIP is developing vector data, it waits in the state of S72 until it can be developed.

(S73):
If any of μRIPa to d can develop the vector data, the command analysis of the tile vector stored in the local memory 19 is performed according to a predetermined grammar.

(S74):
It is determined whether the interpreted command is a drawing command or a paper discharge command. If it is determined that the command is a drawing command, the process branches to S75, and if it is determined that the command is a paper discharge command, the process branches to S76.

(S75):
If it is determined in S74 that the data is a drawing command, a drawing object (DL) is generated. If the command in the tile vector is a character drawing command, a font object is generated based on the font type, character size, and character code specified by the command. If the command is a drawing command other than characters, the command is used. A drawing object of the generated figure (line, circle, polygon, etc.) is generated and stored in the DL area of the local memory 19.

  If it is determined that the print data is not designated by the drawing command, the print position is moved and the print environment is set according to the print data, and the command interpretation for one unit is terminated.

  The above process is repeated until interpretation of all commands in the tile vector is completed.

(S76):
If it is determined that the command is a paper discharge command, μRIP determines whether there is an empty area in the tile raster memory area on the local memory 19. If there is no free space, wait until the other RIP process is completed and free space is available.

(S77):
If there is an empty area in the tile raster memory, the drawing object generated in S75 is read out and drawn (rasterized) in the tile raster area. At this time, if the generation resolution is 600 dpi, the tile raster area is rasterized as a 600 dpi image. Then, the tile raster image that has been drawn is output to the image output unit 15 via the SBB 2.

(S78):
When the command analysis or drawing process for one tile vector is completed in S75 or S77, it is determined whether or not all of the read tile vector memory has been completed. Return and continue processing the next tile vector. If completed, the process proceeds to S79.

(S79):
It is determined whether or not all the processes have been completed for one page of tile vectors. If unprocessed data still remains, the process returns to S71 to read the tile vectors from the system memory and continue the process.

  When all the processes for one page are finished, the tile vector development is finished.

It is a system configuration diagram of the present invention. It is a figure explaining the copy operation | movement flow of the system of this invention. It is a figure explaining the operation flow of printing of the system of the present invention. FIG. 6 is a diagram for explaining an operation flow of transmission of the system in the first exemplary embodiment of the present invention. It is a figure explaining the flow of the raster-> tile vector conversion part of this invention. It is a flowchart which shows the detail of the vectorization process of the drawing other than the character block of this invention, a line | wire, and a table area. It is a flowchart which shows the detail of the vectorization process of the character area of this invention. It is a flowchart which shows the detail of the character vector conversion process of this invention. It is a flowchart which shows the detail of the distortion correction process of this invention. It is a figure explaining the flow of the tile tile page vector conversion part of this invention. It is a figure explaining the flow of RIP (vector expansion | deployment part) of this invention. It is a figure showing the example of the image data handled by this invention. It is a figure showing the example of the page vector data handled by this invention. It is a figure showing the tile area | region of the tile vector handled by this invention. It is a figure showing the example of the tile vector data handled by this invention. It is a figure showing the example of the conventional MFP system. It is explanatory drawing which shows the scan image which the skew produced. It is explanatory drawing which shows the scanned image of a "C" shape.

Claims (1)

In an image distortion correction apparatus that corrects distortion of a scanned image obtained by reading a book document that is in contact with the top or bottom of the contact glass with the page binding portion parallel to the main scanning direction of image reading by the image reading unit,
Image determination means for determining whether the scanned image read by the image reading means is a character image or a photographic image for each block of a predetermined size;
A font information acquisition unit that acquires font information for a block that is determined to be a character image by the image inversion means;
A skew angle detection unit that compares the scanned image with the font information acquired by the font information acquisition unit, and detects a skew angle for each block;
Distortion correction execution means for correcting distortion for each block according to the skew angle detected by the skew angle detection means;
A vector image conversion unit that converts the font information acquired by the font information acquisition unit into a vector image;
An image distortion correction apparatus, comprising: means for expanding a vector image of a block having a predetermined size into raster image data.

Images