JP2004072600A - Image rotation system and image rotation method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve memory access efficiency by burst access without causing a bank conflict and to secure real-time performance during serial scanning. <P>SOLUTION: The image rotation system, which has a receive memory receiving and storing data to be entered and a data buffer temporarily accumulating the data, has means for writing the data in the receive memory, means for reading out the data from the receive memory and writing the data in the data buffer, means for slicing the data from the data buffer and performing image rotation, and means for writing a result after the image rotation into a memory. The system has means for performing burst write access with continuous addresses and switching banks for each line when storing the data in the receive memory, and has means for performing data allocation by performing burst read access switching banks for each line when reading out the data from the receive memory and writing it in the data buffer. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an image data rotation system and an image rotation method for an image forming apparatus including a laser printer, an inkjet printer, and a multifunction peripheral.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, in an image forming apparatus such as a printer, a process of editing image data is required, and for example, a function of rotating image data is also required. It is particularly regarded as important for adjusting the image data to the direction of the paper, and for an ink jet printer that performs serial scanning by arranging ink ejection holes in the vertical direction.
[0003]
An image forming apparatus that performs rotation of image data, particularly raster column conversion, first collects raster data for a certain band and issues an interrupt to the CPU when extracting raster data drawn by some method for each column. The CPU makes various settings for performing raster column conversion and activates the conversion circuit. Thereafter, a part of the raster band data is allocated to a buffer for horizontal h dots and vertical v dots, the data is cut out in the vertical direction and stored in the memory repeatedly. When all the raster band data is processed, the CPU is interrupted. In the order of notification.
[0004]
[Problems to be solved by the invention]
However, the following problems occur when processing this raster column conversion at high speed. FIG. 9 is a diagram showing a relationship between a sheet and an address and a buffer storage diagram when conventional raster sequential addresses are stored in a memory at continuous addresses. In many cases, the format of received data is mainly raster sequential, and is input as an image as shown in FIG. In this case, since the input raster data is continuous, if the data is sequentially written to a continuous address, the data must be read from a discontinuous address and written to a buffer. This is because, when a memory having a bank configuration is used, single access is performed, and the memory access efficiency is extremely deteriorated.
[0005]
Then, when reading into the buffer, a continuous address is set. FIG. 10 is a diagram showing a relationship between a sheet and an address and a buffer storage diagram in the case where conventional raster sequential storage in a memory at discontinuous addresses is performed. In this case, the addresses are arranged in a continuous address order at the stage of packing into the buffer. This means that in the case of burst access, the access is slowed down by the latency of the memory, that is, the read is slowed, so with the intention that the read only manages the burst access and raises the access efficiency even a little. Is also known. However, since the first input is a discontinuous address, that is, a single access, the access efficiency is inevitably deteriorated. However, in a memory access at a discontinuous address, this method may be employed because writing is slightly faster (for latency).
[0006]
Further, when the data is cut out from the buffer in the vertical direction and is sequentially stored in the memory, it is necessary to write the data at a discontinuous address due to the nature of the recording apparatus that performs a serial scan or the like. FIG. 11 is a diagram showing an address image after a conventional buffer is cut out. If writing is performed sequentially without considering the continuity in the vertical direction as shown in FIG. 11, when the head of the ink jet printer operates to read data, the data must be read at a discontinuous address, and real-time performance is ensured. The above is a problem. This is because, in a printing apparatus that performs serial scanning or the like, the number of ejection holes in the vertical direction has been increasing year by year, and it has become impossible to completely fit in the vertical dots of the buffer.
[0007]
Further, when the printing apparatus performs serial scan and sequentially reads the data generated by the column, the required read time is limited, and therefore the data must be read at a continuous address. Although it is devised to write at a discontinuous address, a bank conflict occurs due to the continuous address, and the memory access efficiency deteriorates.
[0008]
The present invention has been made in view of the above circumstances, and has as its object to improve memory access efficiency by burst access without bank conflict, and to ensure real-time performance during serial scan.
[0009]
[Means for Solving the Problems]
In order to achieve this object, the invention according to claim 1 has a reception memory for receiving and storing input data, and a data buffer for temporarily storing the data, and the reception memory has Means for writing data, means for reading the data from the reception memory and writing the data in the data buffer, means for cutting out the data from the data buffer and rotating the image, and writing the result after the image rotation to the memory Means for storing data in the reception memory, performing burst write access at a continuous address and switching a bank for each line, reading data from the reception memory, and storing the data in the data buffer. When writing data to a device, it is necessary to have means for switching the bank for each line, performing burst read access, and arranging data. It is a symptom.
[0010]
The invention according to claim 2 has a receiving memory for receiving and storing input data, a data buffer for temporarily storing the data, means for writing the data in the receiving memory, An image rotation system comprising: means for reading the data from a receiving memory and writing the data in the data buffer; means for cutting out the data from the data buffer to rotate the image; and means for writing the result after the image rotation to the memory. A means for reading the data stored in the data buffer, and a means for switching the bank for each band to write access to the memory and arranging the data when writing the result after image rotation. And
[0011]
The invention according to claim 3 has a receiving memory for receiving and storing input data, a data buffer for temporarily storing the data, a unit for writing the data to the receiving memory, An image rotation system comprising: means for reading the data from a receiving memory and writing the data in the data buffer; means for cutting out the data from the data buffer to rotate the image; and means for writing the result after the image rotation to the memory. Means for reading data stored in the data buffer and writing a result after image rotation, switching a bank for each band, dividing the data into a plurality of address destinations, performing write access, and arranging data. It is characterized by.
[0012]
The invention according to claim 4 has a receiving memory for receiving and storing input data, a data buffer for temporarily storing the data, a means for writing the data to the receiving memory, An image rotation system comprising: means for reading the data from a receiving memory and writing the data in the data buffer; means for cutting out the data from the data buffer to rotate the image; and means for writing the result after the image rotation to the memory. In storing data in the reception memory, there is provided a means for performing burst write access at a continuous address and switching a bank for each line, and when reading from the reception memory and writing to the data buffer, Means to switch the bank for each burst read access and data arrangement, and to write the result after image rotation, And write access to the memory by switching banks, is characterized in that it has means for performing data placement.
[0013]
The invention according to claim 5 has a receiving memory for receiving and storing input data, a data buffer for temporarily storing the data, a unit for writing the data to the receiving memory, An image rotation system comprising: means for reading the data from a receiving memory and writing the data in the data buffer; means for cutting out the data from the data buffer to rotate the image; and means for writing the result after the image rotation to the memory. In storing data in the reception memory, there is provided a means for performing burst write access at a continuous address and switching a bank for each line, and when reading from the reception memory and writing to the data buffer, Means for switching between banks for burst read access and data arrangement, and for storing data stored in the data buffer. When writing the results after image rotation out switches the bank in each band and write access by dividing into a plurality of addressee, is characterized by having a means for performing data placement.
[0014]
The invention according to claim 6 has a receiving memory for receiving and storing input data, a data buffer for temporarily storing the data, and a step of writing the data to the receiving memory; An image rotation method comprising: reading the data from a reception memory and writing the data into the data buffer; cutting out the data from the data buffer to rotate the image; and writing the result after the image rotation to the memory. In storing data in the reception memory, the method includes a step of performing burst write access at a continuous address and switching a bank for each line, and when reading from the reception memory and writing to the data buffer, The method is characterized in that the method includes a step of performing burst read access by switching the bank every time and performing data arrangement.
[0015]
The invention according to claim 7, comprising a receiving memory for receiving and storing input data, and a data buffer for temporarily storing the data, a step of writing the data to the receiving memory, An image rotation method comprising: reading the data from a reception memory and writing the data into the data buffer; cutting out the data from the data buffer to rotate the image; and writing the result after the image rotation to the memory. A step of reading data stored in the data buffer, and a step of writing a result after image rotation, switching a bank for each band, performing a write access to a memory, and performing a data arrangement. Features.
[0016]
The invention according to claim 8 has a receiving memory for receiving and storing input data, a data buffer for temporarily storing the data, and a step of writing the data to the receiving memory; An image rotation method comprising: reading the data from a reception memory and writing the data into the data buffer; cutting out the data from the data buffer to rotate the image; and writing the result after the image rotation to the memory. A step of, when reading data stored in the data buffer and writing a result after image rotation, switching a bank for each band, dividing the data into a plurality of address destinations, performing write access, and performing data arrangement. It is characterized by.
[0017]
The invention according to claim 9 has a receiving memory for receiving and storing input data, a data buffer for temporarily storing the data, and a step of writing the data to the receiving memory; An image rotation method comprising: reading the data from a reception memory and writing the data into the data buffer; cutting out the data from the data buffer to rotate the image; and writing the result after the image rotation to the memory. The step of reading data stored in the data buffer, and the step of storing data in the reception memory by performing a burst write access at a continuous address and switching a bank for each line, When reading from the memory and writing to the data buffer, burst read access is performed by switching banks for each line, and data arrangement is performed. And a step of performing, when writing the results after image rotation, and write access to the memory by switching the bank for each band is characterized by having a step of performing data placement.
[0018]
The invention according to claim 10 has a receiving memory for receiving and storing input data, a data buffer for temporarily storing the data, and a step of writing the data to the receiving memory; An image rotation method comprising: reading the data from a reception memory and writing the data into the data buffer; cutting out the data from the data buffer to rotate the image; and writing the result after the image rotation to the memory. In storing data in the reception memory, the method includes a step of performing burst write access at a continuous address and switching a bank for each line, and when reading from the reception memory and writing to the data buffer, Performing a burst read access by switching banks every time and arranging data; and reading data stored in the data buffer. When writing the results after image rotation switches the bank in each band and write access by dividing into a plurality of addressee, it is characterized by having a step of performing data placement.
[0019]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
[0020]
[Example 1]
FIG. 1 is a block diagram showing one embodiment of the image rotation system of the present invention. The system of the present invention comprises a memory arbiter A1, a memory controller A2, an external memory A3, a DMA block A4, a data receiving section A5, a data source section A6, a data converter A7, a data output block A8, It has a data destination section A9, a CPU A10, and a ROMA11.
[0021]
The memory arbiter A1 performs a process for a memory access request with each functional block. The memory controller A2 manages an interface with the external memory A3. The DMA block A4 manages an interface between each functional block and the memory arbiter A1. The data receiving unit A5 receives some input from the data source unit A6, appropriately switches the address to the transfer destination external memory A3, and transfers the address to the DMA block A4. The data converter A7 performs a rotation process on the data stored in the external memory A3 via the DMA block A4, appropriately switches the address for the transfer destination external memory A3, and transfers the data to the DMA block A4. . The data output block A8 switches the address of the transfer source external memory A3 appropriately, reads the data via the DMA block A4, and outputs the data to some data destination unit A9. The CPU A10 is a CPU that controls the present system. The ROM A11 is a ROM that stores programs and the like.
[0022]
The block diagram in FIG. 1 shows an example of a simple configuration in which input image data is rotated and output. First, the data receiving unit A5 receives the data input by the data source unit A6, and writes the data to the memory A3 via the DMA block A4, the memory arbiter A1, and the memory controller A2.
[0023]
Next, data is input from the memory A3 to the data converter A7 via the memory controller A2, the memory arbiter A1, and the DMA block, and a rotation process is performed (buffer write). The data after the rotation processing is written from the data converter A7 (buffer read) to the memory A3 via the DMA block A4, the memory arbiter A1, and the memory controller A2. The area to be rotated is shifted laterally, and data transmission to the data converter A7 and writing to the memory A3 are repeated, so that the entire area to be rotated is performed.
[0024]
After completion of all the areas to be rotated, the rotated data is read from the memory A3, written to the data destination unit A9 via the memory controller A2, the memory arbiter A1, and the DMA block A4, and the data after the rotation processing is written. Pass out of this process.
[0025]
FIG. 2 is a diagram showing a relationship between paper and addresses and a buffer storage diagram in a case where banks are switched for each raster at consecutive addresses in raster order and stored in a memory. FIG. 2 shows how data input to the data receiving unit A5 in raster order is arranged in the memory A3. Since the memory A3 is a continuous address and the unit is increased from the conventional one (32 bit → 128 bit in FIG. 2), the data is stored in burst units. The bank is switched for each raster, and the data input to the data receiving unit A5 is written to the memory A3.
[0026]
FIG. 3 is a diagram showing a burst write to the reception memory. When filling in the buffer, access is performed in the order of non-consecutive addresses in the order of burst read from address 0 of bank 0, burst read from address 0 of bank 1, burst read from address 0 of bank 2, and stuffing as shown in FIG. Go. Therefore, burst read can be performed without bank conflict. FIG. 4 is a diagram showing burst read at the time of buffer storage. Therefore, according to the present embodiment, bank conflict can be avoided by single access for writing to the memory upon data reception and read for writing into the buffer, and access efficiency can be dramatically improved.
[0027]
FIG. 5 shows a flowchart of the operation of writing data from the data receiving unit A5 to the memory A3. First, the arguments are initialized (step S501), and the receiving raster length H and the storage start address KA are set (step S502). The reception raster length H means how many horizontal units are input with one 128-bit unit. The storage start address KA has a meaning as to which address the start data of the 32 raster data is put. If these two are known in advance, the hardware can calculate and store the address destination for the data input one after another.
[0028]
When 128 bits of data are accumulated (Yes in step S503), the data is stored in the memory (step S504), and the address is added to the destination of 16 bytes (16 * RHT) (step S507). This is repeated until the reception raster length H = RHT, that is, the input of one raster is completed (Yes in step S505). (PHT is the number of received data in units of 128 bits) Switching to the next bank is repeated in the same manner. If a four-bank memory is used, the banks are repeated every four rasters. For every four rasters (Yes in step S508), the address offset is calculated (step S509), and added (step S510). Thus, reception for 32 rasters is performed.
[0029]
FIG. 6 shows a flowchart of an operation of reading data from the memory A3 and storing the data in the buffer in the data converter A7. First, OFFSET in which the address is incremented in units of 128 rasters of 4 rasters, OFFSET2 in which the address is incremented in units of 32 rasters of 128 bits, read and buffer write for 128 * 32 rasters, and incrementing are performed after all processing in the column direction is completed. HV, the address BFA to be written to the buffer after 128-bit read is initialized (step S601). Next, the read start address RA is set, and the read raster length RH is referred to (step S602). The address currently being read is calculated (step S603), read in units of 128 bits, written to the BFA address (step S610), and BFA, R are incremented by 1 (steps S611, S612). This is an access that only changes the bank.
[0030]
When R is a multiple of 4, that is, when reading in 128-bit * 4 raster units is completed (Yes in step S608), 16 * (RH + 1) * (R / 4) is added as OFFSET (step S609), and the BFA address is written. The processing is performed (step S610). In the next 4-raster read in units of 128 bits, addresses for OFFSET are added, and banks are switched.
[0031]
When reading of R = 32, that is, 128-bit * 32 raster units is completed (Yes in step S604), R, OFFSET, and BFA are initialized (step S605), data is sequentially cut out from the column direction (step S606), and HV is increased by one. (Step S613). The point is that the system only writes to the buffer, so we will not mention the specific method. Then, 16 * HV is set as OFFSET2 and added as an offset of the read address (step S614). If RH = HV (Yes in step S607), it means that the target area has been completed, and the process ends.
[0032]
[Example 2]
This embodiment also uses the configurations shown in FIGS. The contents up to this point are other than extracting the data stored in the buffer in the column direction and returning the data to the memory. By the way, the point of the present embodiment is only about the system for writing data from the buffer to the memory. Therefore, the first embodiment in which the system until writing to the buffer is specified is one example, and it is not necessary to particularly specify the system.
[0033]
In a second embodiment, a method of extracting data stored in the buffer in the column direction and returning the data to the memory will be described. The reason why it is described separately from the first embodiment is that the first embodiment and the second embodiment may be used independently or may be used together.
[0034]
FIG. 7 is a diagram illustrating a state of data in a buffer stored in the system according to the first embodiment, a cutout image indicating a vertical cutout order, and an address image after cutout. F127 can be cut out in the order of F0, F1, F2,. In the first embodiment, the bit width of the buffer in the vertical direction is described as 32 bits. However, in this embodiment, the bit width is set to, for example, 128 bits in consideration of burst access in order to further increase the speed. First, by changing to this configuration, the throughput when storing vertically cut data is increased.
[0035]
By the way, the problem is how to write the vertically cut data in the order of F0, F1, F2,..., F127 to the memory A3 in what address order. For example, in the case of an ink jet printer or the like that performs serial scanning by arranging ink ejection holes in the vertical direction, a problem usually occurs. In order to increase the number of nozzle rows in the vertical direction and to operate at high speed, continuous addresses in the vertical direction and to avoid bank conflicts, addresses F0, F1, F2,... And it must be stored by burst access. This is because, by doing so, high-speed serial scanning can be performed, and vertical image data can be read at a continuous address and without bank conflict.
[0036]
The lower part of FIG. 7 is a diagram showing the addresses at which the vertically cut data is arranged. The address offset from F0 to F1 depends on how many addresses are needed in the vertical direction, but in FIG. 7, burst write is performed in a discontinuous order at an offset of 32. By doing so, for example, the head of the ink-jet printer is moved, and certain vertical data is continuously and without bank conflict (in the example, 128 bits from address 0 of bank 0, 128 bits from address 0 of bank 1, 128 bits from bank 0). 2, 128 bits from the address 0 of the bank 2, 128 bits from the address 0 of the bank 3, 128 bits from the address 16 of the bank 3,... (128 bits from the address 16 of the bank 3).
[0037]
[Example 3]
This embodiment also uses the configurations shown in FIGS. The contents up to this point are other than extracting the data stored in the buffer in the column direction and returning the data to the memory. By the way, the point of the present embodiment is only about the system for writing data from the buffer to the memory. Therefore, the first embodiment in which the system until writing to the buffer is specified is one example, and it is not necessary to particularly specify the system.
[0038]
In a third embodiment, a method of extracting data stored in the buffer in the column direction and returning the data to the memory will be described. The reason why the description is made separately from the first embodiment is that the first embodiment and the third embodiment may be used independently or may be used together. In view of these, in this embodiment, the following is performed. The point is that the data stored in the data converter A7 buffer is divided vertically into even holes and odd holes, and then the memory write is performed using the same concept as in the second embodiment. It should be noted that an improvement in access efficiency can be expected by performing only the third embodiment. Further, it may be combined with the first embodiment.
[0039]
FIG. 8 is a diagram illustrating a state of data in a buffer stored in the system according to the first embodiment, a buffer cutout image diagram showing a vertical cutout order, and an EVEN-side address image after the buffer cutout. E127 are extracted in the order of E0, E1, E2... E127, O0, O1, O2. In the first embodiment, the bit width of the buffer in the vertical direction is described as 32 bits. However, in the present embodiment, in order to further increase the speed, it is set to, for example, 256 bits in consideration of burst access. First, by changing to this configuration, the throughput when storing vertically cut data is increased.
[0040]
By the way, the problem is how to write the vertically cut data in the order of E0, E1, E2... E127, O0, O1, O2. For example, in the case of an ink jet printer or the like that performs serial scanning by arranging ink ejection holes in the vertical direction, a problem usually occurs. In order to increase the number of nozzle rows in the vertical direction and to operate at high speed, the addresses of E0, E1, E2... E127, O0, O1, O2. In consideration of this, the data must be stored at discontinuous addresses.
[0041]
The lower part of FIG. 8 is a diagram showing what addresses the vertically cut data is arranged. The number of address offsets from E0 to E1 and from O0 to O1 varies depending on the number required in the vertical direction, but in the figure, burst write is performed in a discontinuous order at an offset of 32. Although the illustration on the ODD side is omitted, it may be arranged in the same manner as on the EVEN side. However, if you put them in exactly the same place, the data written first will of course be lost, so use a different start address.
[0042]
By doing so, for example, while the head of the ink jet printer moves, certain data in the vertical direction is continuous and has no bank conflict (in the example, 128 bits from address 0 of bank 0, 128 bits from address 0 of bank 1 and 128 bits from bank 0). EVEN side data is sequentially read from address 0 of address 2 to 128 bits, address 3 of bank 3 from address 0 to 128 bits, address of bank 0 from address 16 to 128 bits,... For example, 128 bits from address n of bank 0, 128 bits from address n of bank 1, 128 bits from address n of bank 2, 128 bits from address n of bank 3, 128 bits from address n + 16 of bank 0,. Ba It is possible to read the ODD side in the order of 128bit from the address n + address 16 of the click 3.
[0043]
【The invention's effect】
As is apparent from the above description, according to the present invention, the write for storing the input raster data in the memory is stored by burst write, and the read for storing in the buffer is performed in burst without bank conflict. Since data can be stored in the buffer by reading, memory access can be performed at high speed.
[0044]
Further, according to the present invention, when the vertical data is cut out from the buffer and stored in the memory, the vertical buffer is increased and the burst write is performed, and the bank is switched for each band. In the case where real-time performance is required, since the access can be performed without burst read and without bank conflict, memory access can be performed at high speed.
[Brief description of the drawings]
FIG. 1 is a block diagram showing one embodiment of an image rotation system of the present invention.
FIG. 2 is a diagram showing a relationship between a sheet and an address and a buffer storage diagram when a bank is switched for each raster with a continuous address in a raster order and stored in a memory.
FIG. 3 is a diagram showing a burst write to a reception memory.
FIG. 4 is a diagram showing a burst read at the time of buffer storage.
FIG. 5 is a flowchart of an operation of writing data from a data receiving unit A5 to a memory A3.
FIG. 6 is a flowchart of an operation of reading data from a memory A3 and storing the data in a buffer in a data converter A7.
FIG. 7 is a diagram illustrating a state of data in a buffer stored in the system according to the first embodiment, a cutout image indicating an order of vertical cutout, and an address image after cutout.
FIG. 8 is a diagram showing a buffer extraction image diagram showing the state of data in a buffer stored in the system of the first embodiment, the order of vertical extraction, and an EVEN-side address image after the buffer extraction.
FIG. 9 is a diagram illustrating a relationship between a sheet and an address and a buffer storage diagram in the case where conventional raster sequential storage in a memory is performed with continuous addresses.
FIG. 10 is a diagram showing a relationship between a sheet and an address and a buffer storage diagram in the case where conventional raster sequential storage in a memory is performed at discontinuous addresses.
FIG. 11 is a diagram showing a conventional address image after buffer extraction.
[Explanation of symbols]
A1 memory arbiter
A2 Memory controller
A3 external memory
A4 DMA block
A5 Data receiving unit
A6 Data source section
A7 data converter
A8 Data output block
A9 Data destination section
A10 CPU
A11 ROM

Claims (10)

A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Means for writing the data to the receiving memory;
Means for reading the data from the reception memory and writing the data in the data buffer;
Means for cutting out the data from the data buffer and performing image rotation;
Means for writing the result after image rotation to a memory,
When storing data in the reception memory, a burst write access is performed at consecutive addresses, and a unit for switching a bank for each line is provided.
An image rotation system comprising means for performing a burst read access by switching banks for each line and arranging data when reading from the reception memory and writing to the data buffer. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Means for writing the data to the receiving memory;
Means for reading the data from the reception memory and writing the data in the data buffer;
Means for cutting out the data from the data buffer and performing image rotation;
Means for writing the result after image rotation to a memory,
Means for reading data stored in the data buffer,
An image rotation system comprising means for writing data to a memory by switching a bank for each band when writing a result after image rotation, and arranging data. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Means for writing the data to the receiving memory;
Means for reading the data from the reception memory and writing the data in the data buffer;
Means for cutting out the data from the data buffer and performing image rotation;
Means for writing the result after image rotation to a memory,
When reading the data stored in the data buffer and writing the result after rotating the image, a means for switching the bank for each band, dividing the data into a plurality of address destinations, performing write access, and arranging the data is provided. Image rotation system. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Means for writing the data to the receiving memory;
Means for reading the data from the reception memory and writing the data in the data buffer;
Means for cutting out the data from the data buffer and performing image rotation;
Means for writing the result after image rotation to a memory,
When storing data in the reception memory, a burst write access is performed at consecutive addresses, and a unit for switching a bank for each line is provided.
When reading from the receiving memory and writing to the data buffer, means for switching the bank for each line to perform burst read access and arranging data, and for writing the result after image rotation, An image rotation system comprising means for performing write access to a memory by switching between the two and performing data arrangement. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Means for writing the data to the receiving memory;
Means for reading the data from the reception memory and writing the data in the data buffer;
Means for cutting out the data from the data buffer and performing image rotation;
Means for writing the result after image rotation to a memory,
When storing data in the reception memory, a burst write access is performed at consecutive addresses, and a unit for switching a bank for each line is provided.
When reading from the receiving memory and writing to the data buffer, means for performing burst read access by switching banks for each line and performing data arrangement, and reading data stored in the data buffer and performing image read after image rotation An image rotation system comprising means for writing a result, switching a bank for each band, dividing the address into a plurality of addresses, performing write access, and arranging data. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Writing the data in the receiving memory;
Reading the data from the receiving memory and writing the data to the data buffer;
Cutting out the data from the data buffer and performing image rotation;
Writing the result after image rotation to a memory.
When storing data in the receiving memory, the method includes a step of performing burst write access with a continuous address and switching a bank for each line,
An image rotation method for reading data from the reception memory and writing data in the data buffer, the method comprising a step of performing burst read access by switching banks for each line and performing data arrangement. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Writing the data in the receiving memory;
Reading the data from the receiving memory and writing the data to the data buffer;
Cutting out the data from the data buffer and performing image rotation;
Writing the result after image rotation to a memory.
Reading a data stored in the data buffer,
An image rotation method comprising the steps of: writing a result after image rotation, switching a bank for each band, performing write access to a memory, and arranging data. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Writing the data in the receiving memory;
Reading the data from the receiving memory and writing the data to the data buffer;
Cutting out the data from the data buffer and performing image rotation;
Writing the result after image rotation to a memory.
When reading the data stored in the data buffer and writing the result after image rotation, the method includes a step of switching banks for each band, dividing the data into a plurality of address destinations, performing write access, and arranging data. Image rotation method. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Writing the data in the receiving memory;
Reading the data from the receiving memory and writing the data to the data buffer;
Cutting out the data from the data buffer and performing image rotation;
Writing the result after image rotation to a memory.
Reading a data stored in the data buffer,
When storing data in the receiving memory, the method includes a step of performing burst write access with a continuous address and switching a bank for each line,
When reading from the reception memory and writing to the data buffer, the method includes a step of switching the bank for each line, performing burst read access, and arranging data.
An image rotation method comprising the steps of: writing a result after image rotation, switching a bank for each band, performing write access to a memory, and arranging data. A receiving memory for receiving and storing input data;
A data buffer for temporarily storing the data,
Writing the data in the receiving memory;
Reading the data from the receiving memory and writing the data to the data buffer;
Cutting out the data from the data buffer and performing image rotation;
Writing the result after image rotation to a memory.
When storing data in the receiving memory, the method includes a step of performing burst write access with a continuous address and switching a bank for each line,
When reading from the receiving memory and writing to the data buffer, performing a burst read access by switching banks for each line and arranging data;
When reading the data stored in the data buffer and writing the result after image rotation, the method includes a step of switching banks for each band, dividing the data into a plurality of address destinations, performing write access, and arranging data. Image rotation method.

Images