JP2000293678A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To quickly conduct the rotation processing of image data. SOLUTION: This image processor is provided with an image memory 1 which stores image data to be rotated, a block deciding part 2 which divides the image data stored in the image memory 1 into blocks with the number of dots corresponding to the multiple of the number of pixels which can be transferred at once and whose at least one side has the data bus width of a central processing unit, and decides the block to be rotated, a buffer memory 3 which temporarily stores the image data for the number of lines corresponding to the number of dots which can be transferred under one time of data transfer instruction by the central processing unit among the image data of the block decided by the block deciding part 2, and an address controlling part 4 which decides a writing address in a storage memory 5 corresponding to the angle of rotation for the image data for the plural lines stored in the buffer memory 3.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for rotating and outputting image data taken in by a digital copying machine or the like or image data transmitted from a host computer.

[0002]

2. Description of the Related Art A page printer interprets a PDL (page description language) received from a higher-level device (such as a host computer), converts the PDL into raster data corresponding to the resolution of an output device, and prints out the data on paper. I have. Normal,
The drawing object included in the PDL is composed of a drawing command such as a character or a figure, and an image drawing command including a bit string representing a color value and a density value of each pixel. Then, the characters and graphics are converted into raster data having the size, color, and density specified by the dedicated drawing means, respectively, and written at the position specified by the PDL on the output page.

On the other hand, an image drawing command in a PDL generally corresponds to a plurality of image formats having different gradation numbers and resolutions.
Performs image data resolution conversion, color conversion, enlargement / reduction processing, or compression / decompression processing to save the amount of memory used in accordance with the designated value of L and the resolution of the output unit, etc., and is specified by the PDL on the output page. Is written to the specified position.

Further, in the case where the direction of the output paper specified by the PDL and the paper to be actually printed are different, or in the case of a printer having a double-sided printing function, the image is prevented from being inverted on the front and back of the paper. The image rotation processing is performed on raster data or image data to be printed out on the paper as needed.

Conventionally, a process for rotating a large image with a small amount of memory is performed by dividing image data into a plurality of blocks as disclosed in Japanese Patent Application Laid-Open No. 3-37774.
Prepare a buffer memory that can store one line of image data, read the image data to be rotated into the buffer memory on a line-by-line basis, calculate the write address from the position and rotation angle of the destination block, and read the image from the buffer memory. This is realized by repeating the process of writing data to the memory that stores data. At this time, since the reading and writing processes of the image data are all performed for each pixel, the number of times of reading and writing the image data required to rotate the entire image is equal to the total number of pixels of the image.

[0006]

However, in recent years, the amount of image data has been remarkably increased due to the progress of higher resolution and colorization of printers, and this has caused not only an increase in the amount of memory required for storing images, but also an increase in MPU. Also, an increase in the image rotation processing time due to an increase in the number of data transfer instructions executed by the computer has become a serious problem. In particular, since many MPUs have a built-in cache mechanism these days, if the number of data transfer instructions increases, cache flushing frequently occurs implicitly inside the MPU, causing a reduction in the processing speed of the printer device. Become.

[0007]

SUMMARY OF THE INVENTION The present invention is an image processing apparatus designed to solve such a problem. That is, the image processing apparatus of the present invention includes an image memory for storing image data to be rotated, and the number of pixels at least one side of which can transfer image data stored in the image memory at a time with the data bus width of the central processing unit. Block deciding means for dividing into blocks having the number of dots corresponding to a multiple of, and determining a block to be rotated, and one-time data transfer by the central processing unit among image data of the blocks decided by the block deciding means Determines buffer memory for temporarily storing image data for the number of lines corresponding to the number of dots that can be transferred by the instruction, and storage memory write address according to the rotation angle for image data for multiple lines stored in the buffer memory Address control means.

According to the present invention, when rotating image data, the number of dots corresponding to a multiple of the number of pixels that can be transferred at a time with at least one side of the image data to be rotated with the data bus width of the central processing unit. , And the block to be rotated is determined by the block determining means, stored in the image memory, and then stored in a rotated state in the storage memory by the address control of the address control means. By preparing a plurality of lines corresponding to the number of dots that can be transferred by one data transfer instruction by the central processing unit corresponding to the block width as the buffer memory, data to the buffer memory corresponding to the data bus width is prepared. In addition to storing data, the data of the number of dots that can be transferred from the buffer memory by a single data transfer instruction by the central processing unit can be packed and sequentially written to the storage memory. Reading and writing can be performed.

[0009]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of an image processing apparatus according to the present invention. FIG. 1 is a block diagram illustrating an image processing apparatus according to the present embodiment. That is, the image processing apparatus includes an image memory 1 for storing input image data, a block determining unit 2 for dividing the image data stored in the image memory 1 into a plurality of blocks, and determining a block to be rotated. A buffer memory 3 for temporarily storing a plurality of lines of the image data of the block to be rotated and a write address to the storage memory 5 corresponding to the rotation angle for the image data temporarily stored in the buffer memory 3 are determined. And an address control unit 4.

The image memory 1 temporarily stores image data captured by a scanner and image data transmitted from a host computer via a network.

The block determining unit 2 determines the number of pixels that can transfer image data stored in the image memory 1 at a time at least on one side with a data bus width of a central processing unit (hereinafter, referred to as “MPU”). This is a part that divides into blocks having the number of dots (number of pixels) corresponding to a multiple of, and determines the block to be rotated. For example, MP
When the data bus width of U is 16 bits and the number of bits per pixel is 4 bits, a square area that is a multiple of 4 is defined as one block. Note that the block may be a rectangular area. As a result, the data bus can be used effectively when transferring image data from the image memory 1 to the buffer memory 3.

The buffer memory 3 stores the M data of the image data of the block determined by the block determination unit 2.
The image data for the number of lines corresponding to the number of dots (the number of pixels) that can be transferred by one data transfer command by the PU is temporarily stored.

The address control section 4 controls a read address and a write address to the storage memory 5 for a plurality of lines of image data stored in the buffer memory 3. By this address control, the image data read from the buffer memory 3 is stored in the storage memory 5 while being rotated by a predetermined angle.

FIG. 2 is a flowchart illustrating the flow of an image rotation process in the image processing apparatus according to the present embodiment. In the present embodiment, rotation processing is performed by dividing image data of an arbitrary size stored in the image memory 1 (see FIG. 1) into a plurality of blocks each having the same width and height and each having an aligned size. Therefore, first, in step S101, the number of block divisions in the row and column directions of the image data stored in the image memory 1 is obtained.

Here, (the height of the block × the number of block divisions in the column direction) and (the width of the block × the number of block divisions in the row direction) are equal to the height and width of the image data to be rotated, respectively. Or larger,
As shown in FIG. 3, a blank portion without image data may be formed around a block.

In step S102, the number of pixels in the margin of each block is set as an offset value from the block boundary to the image data portion so that each pixel can be identified as image data or a margin in a later process. Acquire (see FIG. 3).

Next, in step S103, the buffer memory 3 having the same width as the block and having the same number of pixels as the number of pixels that can be transferred by one data transfer instruction in the MPU to temporarily store the rotating image data. (Figure 1
See). Subsequently, in step S104, as an area for storing the image data after the rotation processing,
A storage memory 5 (see FIG. 1) having a size equal to (the number of pixels per block × the number of block divisions in the column direction × the number of block divisions in the row direction) is secured.

Here, the start address of this area is MPU
The data bus width must be aligned. Next, in the loop processing of steps S105 to S112, rotation processing of image data is performed on each block. Here, as the moving order of the block to be subjected to the rotation processing, the uppermost row and the leftmost column shown in FIG.
An example is a case in which the rule moves to the right in the row direction and returns to the leftmost position in the next column when the rightmost position in a certain row is reached.

First, in step S105, the block to be processed is set in the uppermost row and the leftmost column. Next, in step S106, a block to which a block to be processed is pasted in the rotated image is determined based on the rotation angle. 4A and 4B are diagrams for explaining a block after rotation. FIG. 4A shows a state before rotation, and FIG.
(C) shows a state rotated 90 degrees clockwise. That is, when the block in the uppermost row and the leftmost column shown in FIG. 4A is rotated counterclockwise by 90 degrees, the block is pasted to the block in the lowermost row and the leftmost column as shown in FIG. To rotate 90 degrees to (c)
As shown in the above, it is pasted to the block in the top row and right column.

Next, in step S107, the image data of the block to be processed is temporarily transferred to the buffer memory 3 (see FIG. 1) secured in step 103, and then rotated, and the storage memory 5 secured in step S104 (see FIG. 1). 1), a process of pasting to the block of the paste destination determined in step S106 is performed. Details of the processing of this part will be described later.

Subsequently, in step S108, it is determined whether the block to be processed is at the right end of the row. If not, the block to be processed is moved right by one column in step S109, and the process returns to step S106. In step S110, it is determined whether or not the block is the bottom row. If not, the block to be processed is moved to the leftmost column one row below in step S111 and step S112.
Return to If it is the bottom line, the rotation processing of all blocks has been completed, and the processing exits from the loop processing. At this point, the rotated image data is stored in the memory secured in step S104.

Next, the details of the block rotation / pasting processing in step S107 will be described with reference to the flowchart of FIG. This processing is performed in the image memory 1
(See FIG. 1) The image data of the block stored above is rotated at a designated angle while being temporarily transferred to the buffer memory 3 (see FIG. 1), and the image data after the rotation processing is stored in the storage memory 5 (see FIG. 1). ) Is attached to the area of the specified block.

Here, the rotation within a certain block
The order in which the pixels to be pasted are moved
An example will be described in which the rule is such that the top row and the leftmost column are the first, the row is moved to the right in order in the row direction, and if it reaches the rightmost position in a row, it returns to the leftmost position in the next column.

First, in step S201, a pixel to be processed first is set in the uppermost row and the leftmost column. Next, in step S202, the number-of-pastes counter is cleared to zero. The value of this counter will be referred to when determining the pasting position later. Subsequently, the pixel to be processed in step S203 is set in step S103 described above (see FIG. 2).
The writing position for transferring the data to the buffer memory 3 (see FIG. 1) secured in step (1) is set to the uppermost row and the leftmost column.

Next, it is determined in step S204 whether or not the pixel to be processed is a margin using the offset value obtained in step S102 (see FIG. 2). Is transferred to the write position on the buffer memory 3 (see FIG. 1), and in the case of a margin, data in which the color value of each color is set to 0 is transferred as pseudo data in step S206.

Subsequently, in step S207, it is determined whether or not the pixel to be processed is at the right end of the row. If not, the pixel to be processed and the write position of the buffer memory 3 (see FIG. 1) are set to one column in step S208. It moves to the right and returns to step S204. If it is the right end, it is determined in step S209 whether or not the buffer memory 3 is full.
If it is not full, in step S210 and step S211, the pixel to be processed and the writing position of the buffer memory 3 are moved to the leftmost column one row below and the process returns to step S204.

If the buffer memory 3 (see FIG. 1) is full, the flow advances to step S212 to rotate the contents of the buffer memory 3 while rotating the contents of the storage memory 5 (see FIG. 1) in step S104 (see FIG. 2). S106 (FIG. 2
(See Reference)). Details of the processing of this part will be described later.

When the pasting process is completed, step S2
In step 13, the number of times of pasting is increased by one. Subsequently, in step S214, it is determined whether the pixel to be processed is the bottom row. If not, the pixel to be processed is moved to the leftmost column one row down in steps S215 and S216, and the process returns to step S203. . On the other hand, in the case of the bottom line, the rotation processing of all the pixels in the block has been completed.
Exit from loop processing. At this time, the image data of the rotated block to be processed is stored in the area of the designated block on the storage memory 5 (see FIG. 1) secured in step S104 (see FIG. 2).

Next, the details of the data pasting process in step S212 will be described with reference to the flowchart of FIG. In this process, the pixel data transferred to the buffer memory 3 (see FIG. 1) is efficiently attached to the designated block of the storage memory 5 (see FIG. 1) while rotating at the designated angle.

First, in step S301, the destination address of the pixel data on the buffer memory 3 (see FIG. 1) on the storage memory 5 (see FIG. 1) is determined.
This is performed based on the rotation angle, the block to be pasted specified in step S106 (see FIG. 2), and the value of the above-mentioned pasting frequency counter.

FIG. 7 is a schematic diagram for explaining the data pasting process, in which the data is rotated by 90 degrees counterclockwise. This block is a square area of 16 pixels × 16 pixels, and the buffer memory 3 is composed of a block width (16 pixels) × the number of pixels (4 pixels) that can be transferred by one data transfer command of MPU. I have. Therefore, the sticking frequency counter counts up from 0 to 3.

The address to which the image data once stored in the buffer memory 3 is to be pasted to the storage memory is within the range of the pixel to which the block of the storage memory is to be pasted, which is determined by the pasting counter and the rotation angle. Become.

Next, steps S302 to S30
By the loop process 7, the pixels having the same row direction position of each line in the buffer memory 3 (see FIG. 7) are packed as a group of data, and the rotated image is pasted on the storage memory 5.

That is, in step S302, the position of the pixel to be pasted first in the buffer in the row direction is set to the leftmost position. Next, in step S303, pixels in the row direction to be pasted are read from each row in the buffer memory 3 (see FIG. 7), and are arranged in the data register of the MPU in an order according to the rotation angle.

In securing the buffer memory 3 (see FIG. 7) in step S103 (see FIG. 2), the number of lines is the same as the number of pixels that can be transferred by one data transfer instruction in the MPU. The pixel data for all lines of the memory 3 just fills the data register of the MPU.

Subsequently, in step S304, the contents of the register are transferred to the paste destination address of the rotated image storage memory. This transfer command transfers a plurality of pixel data in one execution. Next, in step S305, the row direction position of the pixel to be pasted in the buffer is shifted right by one, and in step S307, the next pasting destination address is determined.

If it is at the far right, it means that all the data in the buffer has been pasted, and the processing exits from the loop processing.

FIG. 8 is a schematic diagram for explaining the rotational attachment. That is, in the image data read from the rotation source block and stored in the buffer memory 2, four pixels (A to D) in the same row are read sequentially from the leftmost row. When the read four-pixel data is rotated 90 degrees in the counterclockwise direction, the data is written in the order of (A, B, C, D) from the bottom row of the block to which the data is pasted, and clockwise. When the image is rotated 90 degrees, data is written in the order of (D, C, B, A) from the top row of the block to be pasted.

FIG. 9 is a schematic diagram for explaining a sticking position. After writing 4 pixels to the paste destination block as described above, move the 4 pixels to be pasted to the right by 1
If the pixel is rotated 90 degrees counterclockwise, it is written in the row immediately above the previously written 4 pixels, and if it is rotated 90 degrees clockwise, 1 row of the previously written 4 pixels is rotated. Write to the line below. By repeating this sequentially, it becomes possible to write all the image data stored in the buffer memory 2 in the block of the paste destination in a rotated state.

After the rotation pasting process for one block is completed, the same rotation pasting process is repeated for the next block. When the rotation pasting process for all the blocks is completed, the rotated data of all the image memories is stored in the storage memory 5 (see FIG. 1).

In the present embodiment, when a pixel to be subjected to the rotation process in a certain block is a margin, the data is included in the block as pseudo data instead of using data with the color value of each color being 0. It is also possible to use the pixel data of the outermost peripheral part of the image data to be used.
This makes it possible to improve the compression ratio when performing compression in the subsequent processing.

[0042]

As described above, the image processing apparatus according to the present invention has the following effects. That is, since the data transfer from the buffer memory to the storage memory is performed simultaneously by a single transfer command for a plurality of pixels according to the data bus width, the number of executions of the data transfer command can be greatly reduced. It is possible to optimize cache access that occurs inside the central processing unit. This makes it possible to significantly speed up the rotation processing of large image data.

Further, when there is a margin of image data in a block, by assigning pseudo data to the margin portion, it is possible to speed up the rotation processing of the image data of the block. Further, by using the pixel data of the outermost peripheral portion of the image data included in the block in the blank portion, it is possible to improve the compression ratio when compressing the image in the subsequent processing, thereby reducing the amount of memory. It becomes possible.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an image processing apparatus according to an embodiment.

FIG. 2 is a flowchart illustrating a flow of an image rotation process.

FIG. 3 is a schematic diagram illustrating a blank portion in a block.

FIG. 4 is a diagram illustrating a block after rotation.

FIG. 5 is a flowchart illustrating a rotation / pasting process.

FIG. 6 is a flowchart illustrating a pasting process.

FIG. 7 is a schematic diagram illustrating a data pasting process.

FIG. 8 is a schematic diagram for explaining the rotational attachment.

FIG. 9 is a schematic diagram for explaining a sticking position.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Image memory, 2 ... Block determination part, 3 ... Buffer memory, 4 ... Address control part, 5 ... Storage memory

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2C087 AB05 BC02 BC05 BC06 BD24 5B047 EA07 EB11 5B057 CA12 CA16 CB12 CB16 CD04 CH11 9A001 BB03 BB04 EE02 HH34 JJ35 KK42

Claims (3)

[Claims] 1. An image memory for storing image data to be rotated, and at least one side of the image data stored in the image memory corresponds to a multiple of the number of pixels that can be transferred at a time with a data bus width of a central processing unit. Block deciding means for dividing into blocks having the same number of dots and determining a block to be rotated, and transferring the image data of the block decided by the block deciding means by one data transfer instruction by the central processing unit. A buffer memory for temporarily storing image data for a number of lines corresponding to a possible number of dots; and an address for determining a storage memory write address corresponding to a rotation angle for the image data for a plurality of lines stored in the buffer memory. An image processing apparatus comprising: a control unit. 2. The image processing apparatus according to claim 1, wherein when there is a dot having no image data among the dots constituting the block determined by the block determination means, pseudo data is assigned to the dot. 3. The image processing apparatus according to claim 2, wherein the pseudo data comprises outermost data in image data of a block having a dot without image data.