JPH05902B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for storing color image information, and more particularly to a method for storing color-specific image information obtained by color separation.

2. Description of the Related Art Color image processing apparatuses have been proposed that separate and read a color image, transmit the read signal, or record a color image based on the read signal.

For example, when color-separating a color original such as a book or photograph to read and reproduce the colors, the original image is first optically divided into blue (B), green (G), and red (R) using a filter, dichroic mirror, etc. 3 of light
It is separated into primary colors and each is converted into an electrical signal by photoelectric conversion means such as a CCD line sensor. Based on the signals of the three primary colors of light, yellow (Y),
An image signal indicating the three primary colors of magenta (M) and cyan (C), or these three primary colors and black (BK) is formed. In accordance with the image signal obtained in this way, full-color image reproduction is performed by superimposing the images of each color and performing a recording operation.

Obtaining the three primary colors, including black as necessary, from the three primary colors of light described above is referred to as color conversion herein. The simplest method of color conversion is blue light (hereinafter referred to as B).
) is converted into a yellow dye (hereinafter referred to as Y) using a negative correlation, and in the same way, green light (hereinafter referred to as G) is converted into a magenta dye (hereinafter referred to as M), and red light (hereinafter referred to as M) is converted into yellow dye (hereinafter referred to as Y). There is a dye that converts a cyan dye (hereinafter referred to as R) into a cyan dye (hereinafter referred to as C). However, the accuracy of color separation of B, G, R and Y, M,
The purity of ink, etc. of C does not completely match the theoretical value, and simply using a negative correlation cannot be said to be an optimal method. Therefore, color correction processing such as masking, which is well known in color printing methods, is required.

For example, when reading a certain section on a document by separating it into three colors, B, G, and R, and forming image signals corresponding to Y, M, and C accordingly, these values are I _B and I, respectively. Let _G , I _R , O _Y , O _M , O _C be constants
Using a ₁₁ to _{a 33} , O _Y = a ₁₁ I _B + a ₁₂ I _G + a ₁₃ I _R O _M = a ₂₁ I _B + a ₂₂ I _G + a ₂₃ I _R O _C = a ₃₁ I _B + a ₃₂ I _G A color conversion process according to a formula such as +a ₃₃ I _R is generally used.

As described above, color image information processing such as color conversion often requires a plurality of color signals obtained by color-separating the same section in order to obtain signal values for the target section.

FIG. 1 shows the general configuration of a color image processing device. 1 is for color images such as originals in B, G, and R.
A reader unit has a line sensor such as a CCD in which multiple photoelectric conversion elements are arranged in a row for photoelectric conversion for each color, and 2 is a reader unit that serially converts multiple pixels in line units for B, G, and R from the reader unit 1. An image memory that divides the output image information into units of a predetermined number of pixels (for example, 16 pixels) and stores them in parallel.
In order to store B, G, and R signals obtained from at least one page of a document, a semiconductor memory having a storage capacity of three pages is provided. 4 is a control unit that has an arithmetic circuit such as a microcomputer, and performs image processing such as the color conversion described above on the image information stored in the image memory 2; , M, and C, a printer unit records a color image on a recording material such as paper based on the signals for 5, a bus for transferring control information and image information, 6a,
6b is an address line, 7a and 7b are data lines, 8a and 8b are control lines, and 9 is a bank line for designating an upper address of the memory from the control unit 4 to the image memory 2. The reader section 1, image memory 2, reader section 3, and control section 4 shown in the first figure may be housed in one housing, or may be divided into several units.

The operation of the color image processing apparatus shown in FIG. 1 will be explained.

To simplify the explanation, 297mm x
The processing of image information for a 208 mm (approximately A4 size) document will be described below. The reader section 1 has a reading capacity of 16 pixels per 1 mm in the main scanning direction, and 1 in the sub scanning direction.
Assuming that 16 lines are scanned per mm, the original is converted into a digital signal indicating 4752 bits of brightness and darkness in one scan. Further, sub-scanning is performed for 3328 lines, and each color is read as a pixel of 15814656 bits. If data representing brightness and darkness is stored in the image memory 2 in correspondence with one bit each, a memory area of 1976832 bytes is required for each color. Therefore, when reading, B, G,
5930496 bytes for R3 colors, 7907328 bytes are required to hold each data of Y, M, C obtained by color conversion processing and black pixel (hereinafter referred to as BK) obtained by inking processing. shall be.

On the other hand, if the computer included in the control unit 4 is an Intel iSBC86/12, and the bus 5 is an Intel Multibuzz, the address space of the computer is 20 bits, so
The maximum amount of accessible memory is limited to 1048576 bytes (1MB). Generally, in order to access a large capacity memory by a computer with such a small address space, a technique is often used to expand the addresses accessible by the computer by switching banks.

FIG. 2 shows a conceptual diagram of bank switching. 10 is the address space of the computer, and in the case of a 20-bit address space, it is 00000 to FFFFF in hexadecimal notation.
It becomes a space of In the following explanation, hexadecimal notation will be expressed as FFFFF ^H by adding an H to the smallest digit. In this space 10, the address space 10 is divided into two, and a space 10a from 00000 ^H to 7FFFF ^H is prepared as a bank switching target space. At this time, image memory 2 is the address space
If it has a storage capacity of 8MB (8388608 bytes) with ^000000H to ^7FFFFFH , divide it into 16 parts and store ^{000000H to} 07FFFFH in the 0th bank B0,
080000 ^H ~0FFFFF ^H to 1st bank B1, 780000 ^H
~ ^7FFFFFH is defined as the 15th bank B15. Then, from the control unit 4 to the image memory 2 via the bank line 9, the data is sent from the 0th bank B0 to the 15th bank B15.
When a code corresponding to any of the above is output, the target bank can be used as the space 10a.
For example, the bank code corresponding to the first bank B1 is
When it is 1 ^H , when the control section 4 outputs 1 ^H to the bank line 9 and refers to the memory at 03000 ^H , the control section 4 actually refers to the contents at 083000 ^H in the image memory 2. As shown in the address conversion diagram in Figure 3, this corresponds to the 23-bit physical address of image memory 2.
This can be achieved by generating RA using a 4-bit signal BA for bank switching and 19 bits of a 20-bit multibus address MA.

FIG. 4 shows a memory map when the image information read from the reader unit 1 after color-separating the original into three colors, B, G, and R, is written into the image memory 2 under the above conditions. Further, the order in which image information is stored in the image memory 2 is shown in FIG. In this way, three-color or four-color image information obtained from the same document area is stored in memories in different banks. In addition, 16 bits of image information serially input from the reader section 1 for each line are collected, each of which is treated as one word, and an address is attached to each 2-byte data that makes up one word. Give and store 32 discrete addresses. By doing this, it is possible to store images in sections of 1 mm x 1 mm in the memory area at consecutive addresses,
This makes it easier to perform later image processing.

This storage method is detailed in Japanese Patent Application Laid-Open No. 78253/1983, so please refer to it.

In the above manner, B and B are stored in image memory 2.
After the image information for each color of G and R is stored, this information is used to convert into image information for each color of Y, M, C, and BK. This procedure is shown in FIG. By steps 601, 602, and 603, zero in the sub-scanning direction is set.
Construct a loop that sequentially processes the area divided into 208 areas from mm to 208 mm, and step 604, 605, 606
A loop is formed to sequentially process 297 areas from 0 mm to 297 mm in the main scanning direction. Therefore, steps 607, 608, and 609 are processes for sections of 1 mm x 1 mm, but this corresponds to one page of the manuscript, i.e.
Repeated for 61776 sections. By the way, step
When handling data in image memory 2 with 607, 608, and 609, as mentioned above, for bank line 9,
The bank code for bank switching must be output separately from the address data. Repeat this step in step 7
As shown in the figure. That is, when the address of a word in the image memory 2 that stores the image information to be accessed is a, in step 701, a bank code x and an intra-bank address y are calculated from the address a.
For example, when using the address conversion shown in Figure 3, the most significant 4 bits of address a are converted to bank code x.
Let 0 be added to the beginning of the lower 19 bits of a to make 20 bits, and this is the in-bank address y. In step 702, bank code x is output to bank line 9, and in step 3, image memory 2 is accessed using address y in the bank as a memory address.

In this way, the sixth illustrated step 607,
In steps 608 and 609, the processing shown in FIG. 7 is required every time image data on the image memory 2 is handled, and bank switching processing must be performed frequently. Furthermore, this corresponds to steps 601, 602, 603, 604, 605, and
Since the loop consisting of 606 repeats 61776 times, a large amount of time is wasted just for memory access.

In view of the above points, the present invention reduces the frequency of memory bank switching processing even when performing color correction such as combining color image information of multiple color components.
An object of the present invention is to provide a color image information storage method suitable for high-speed reading of color image data.

FIG. 8 is a memory map showing an example of the order in which image information is stored in the image memory of a color image processing apparatus to which the present invention is applied. Since the configuration of the color image processing apparatus in the present invention does not need to be changed from the conventional one, reference is made to FIG.

As shown in FIG. 8, image areas for Y, M, C, and BK are sequentially and repeatedly stored by 1 mm in the pixel sub-scanning direction. In the description of the present invention, similarly to the prior art described above, the image memory 2 has addresses ^{000000H to} 7FFFFFH, and this is divided into 16 banks B0 to B15, and access to the image memory 2 is The address space of the computer is expanded by using a 20-bit multi-bus address and a 4-bit bank switching signal. Further, one line of serial signals inputted from the reader section 1 is stored with an address assigned every 16 bits.

At this time, in the 0th bank B0, that is, an area of 524288 bytes from the beginning of the image memory 2, image information for each color of B, G, and R between 0 mm and 13 mm in the sub-scanning direction is stored for 1 mm in the sub-scanning direction. Main scanning direction 297mm
Repeat for each minute block. However, from each image information of B, G, R, Y, M, C,
When reading the image information from the reader unit 1 to the image memory 2 to create image information of four colors of BK, B,
Dummy blocks of the same size as B, G, and R are prepared between the G and R blocks and the next B, G, and R blocks, and nothing is read into the dummy blocks. FIG. 9 shows the address map of the image memory 2 when the image information for each color of one page of the original is read in this way. 9th
As is clear from the figure, data with a width of 13 mm in the sub-scanning direction is assigned to each bank.

The image processing procedure when doing this is explained in the 10th section.
As shown in the flowchart in Figure. step 1001,
1002, 1003, and 1004 form a loop in which image processing is performed for each bank. Also, in steps 1005, 1006, and 1007, image processing in the sub-scanning direction in each bank is repeated 13 times in 1 mm units.
Steps 1008, 1009, and 1010 constitute a loop such that image processing in the main scanning direction for 1 mm in each sub-scanning direction in each bank is performed 297 times in units of 1 mm. Therefore, steps 1011, 1012,
The image processing for the image information of 1013 1 mm x 1 mm sections is repeated 61776 times for the entire document.

In this way, when processing a 1 mm x 1 mm image, the number of accesses to image memory 2 can be reduced.
Compared to the conventional method which requires overhead related to extended addresses such as several bank switches multiplied by 61776 times, using the method of this embodiment, bank switching only needs to be performed 16 times in step 1003. Furthermore, steps 1011, 1012,
When handling image memory 2 with 1013, all addresses within the bank can be used, so
Address conversion becomes easy and processing can be significantly speeded up.

In order to explain this embodiment, Japanese Patent Laid-Open No. 57
Although the method of storing image data in the image memory introduced in Publication No. 78253 was used, other methods, such as combining image data input line by line from the reader unit 1 into one word of 16 bits each, into an image memory with consecutive addresses. It may be stored in area 2. Furthermore, it is clear that even if B, G, and R are not stored alternately in that order, they may be rearranged so that they are in the same unit in the extended address space, that is, in the same bank in the above example. be.

In addition, although we have discussed a method in which each input point of each color is expressed using 2 bits to represent brightness or darkness, it goes without saying that this method can also be applied to a method in which multiple bits are input per point to represent the depth, or brightness, of a single point. .

Further, in addition to image information read from a document, the present invention can also be applied to a device that outputs an electrical signal having color information such as a video camera.

Furthermore, as the image memory 2, a magnetic or optical disk may be used in addition to the semiconductor memory.
In this case, since the memory capacity is even larger, the effect becomes more significant.

As explained above, according to the present invention, color image information of multiple color components belonging to the same area of a color image is stored in the same bank, so that even when reading multiple color components from such memory and performing color correction, There is no need to frequently switch banks, and color information can be read out from the storage means at high speed.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an example of the configuration of a color image processing device, Fig. 2 is a conceptual diagram showing bank switching, Fig. 3 is a conceptual diagram showing address conversion for address expansion, and Fig. 4 is a conceptual diagram showing a conventional color image processing device. A memory map diagram showing an example of storing image information in the image memory of a color image processing device. FIG. 5 is a memory map diagram showing the storage order of image information in a conventional color image processing device. FIG. 7th flowchart showing image processing in the processing device
FIG. 8 is a flowchart showing an image memory access procedure in image processing in a conventional color image processing device, and FIG. 8 is a memory map showing an example of the storage order of image information in the color image processing device of the present invention. FIG. 9 is a memory map diagram showing an example of storing image information in the image memory of the color image processing device of the present invention, and FIG. 10 is a flowchart diagram showing image processing in the color image processing device of the present invention. 1 is a reader section, 2 is an image memory, 3 is a printer section, 4 is a control section, 5 is a bus,
6a, 6b are bus address lines, 9 is a bank line, 10 is an address space, B0, B1, B15
is a bank of image memory.

Claims (1)

[Claims] 1. When storing color image information of a plurality of color components representing a predetermined color image in a color image memory having a plurality of banks, storing color image information of the plurality of color components belonging to the same area of the color image in the same bank. A color image information storage method characterized by: