JPH09224134A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent a dead lock while preventing the dispersion of space areas on a buffer and minimizing image data waiting the space area. SOLUTION: At the time of writing next image data in a page buffer (ring buffer) 10, an effective space area capable of storing the next space area is searched. The search is started from a writing pointer P and when the effective space area is not found in spite of reaching a prescribed search finishing condition, the searching is finished to maintain the state of waiting writing. As the search finishing condition, the number of skips from a space area to the next space area is used. As the probability that each space area after the reading of image data is continued is heightened, the dead lock is prevented.

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 having a buffer for temporarily storing image data,
For efficient buffer management.

[0002]

2. Description of the Related Art In recent years, in image processing apparatuses such as digital copying machines, printers, facsimiles, and multi-function machines combining these functions, generally, the speed difference between the image input section and the image output section is reduced. In order to absorb, an image storage unit is mounted in the device to some extent. As the image storage unit, a hard disk or a magneto-optical disk that can store a large amount of data at a low price but has a slow write / read speed, and a semiconductor memory that is expensive but has a fast write / read speed are known. . Image processing apparatuses in recent years use them in combination in order to make the best use of their characteristics. By mounting such two types of image storage units, not only the speed difference absorption described above but also various and effective image processing can be realized. That is, when such two types of image storage units are used together, for example, the function of storing the input image data in the image storage unit and outputting it a plurality of times, the image data in an output order different from the input order It is possible to realize functions that are indispensable for recent image processing apparatuses, such as an electronic sorting function for outputting the.

On the other hand, a method of compressing image data has become widespread in order to reduce the capacity of the image storage section and to transfer the image data at high speed. By compressing the image data, a large number of pages of images can be stored at one time, and the transfer time such as writing / reading can be shortened, so that high-speed processing becomes possible. However, according to a general variable length compression process, the compression rate varies depending on the content of the image data, so that the amount of data after compression varies for each plate. The plate corresponds to the image of each page in the case of a monochrome image, and corresponds to the image of each page for each color in the case of a color image.

Therefore, in order to miniaturize and efficiently use a semiconductor memory (hereinafter referred to as a page buffer) capable of storing a plurality of image data, the buffer next to the last address is set to the first address to make the buffer virtual. A device that uses a page buffer (hereinafter referred to as a ring buffer) that is circular in nature is disclosed in, for example, Japanese Patent Laid-Open No. 60-100872. In the system which realizes the above electronic sorting function by using the page buffer and the hard disk while performing the compression process, the ring buffer method can be said to be the most suitable method at present.

However, in the image processing apparatus, the image input order and the image output order are not always the same due to the double-sided / single-sided processing, the number of print copies, the operation sequence of the output apparatus, and the like. For example, for an output device capable of double-sided output (for example, a printer), a device that outputs in the order of front, back, front, ..., or a device that outputs in the order of front, front, front, back, back, back, ... Exists. Therefore, even if the input order is adjusted to a specific output device, it is impossible to output to the other output device in the input order. In this way, when the input order and the output order are different, in the above-mentioned ring buffer method that handles compressed images, free areas in the buffer are scattered when areas where compressed images are stored are released in a predetermined order. Will end up.

Since the empty area is an area in which the image data after variable length compression is stored, the size of the area varies, and even if there is an empty area, the next compressed image data is not always present. Can not be stored. That is, the ring buffer is characterized by being able to secure a continuous free area, but due to the difference in the input order and the output order, it cannot exhibit its characteristics. In some cases, the buffer as a whole has a sufficient free area. There is a problem (deadlock) in which the next image data cannot be stored due to the dispersion of each free area,
There was a problem that input processing and output processing could not be performed in parallel.

In this regard, as disclosed in Japanese Patent Laid-Open No. 2-288464, the buffer storage order is changed according to the page output order (descending order), and two-sided output is possible as in Japanese Patent Laid-Open No. 60-74768. In which a buffer is divided into two parts, one for the front side and the other for the back side, and stored.
When deadlock occurs in the buffer due to a different order as in Japanese Patent Publication, a page different from the output order (page convenient for releasing the buffer area) is output to avoid the deadlock. Various methods such as things have been proposed.

[0008]

However, the buffer management system as disclosed in Japanese Patent Laid-Open No. 60-74768 described above only supports double-sided output, and output by double-sided / single-sided switching. The change in order and the efficient management of the buffer in the case of outputting a plurality of times are not sufficiently considered.

Further, in the deadlock avoidance method disclosed in Japanese Patent Laid-Open No. 6-86030, since the output order is different from the output order intended by the user, the output sheets are rearranged. There is a problem that it has to be done manually and is troublesome.

Further, in the buffer storage method disclosed in Japanese Patent Laid-Open No. 2-288464, since the image data is stored in the buffer in the output order, there is no problem such as deadlock, but the same image is generated. Since the number of data outputs is not considered, for example, when outputting only specific image data multiple times, the buffer area is released by the first output and the second output is attempted. To read the image data from the hard disk again, as a result,
The processing speed was slow.

Further, in the buffer management method which is a known technique and divides a buffer into block units and is used, even if the input order and the output order are different, since the buffer is cut out in page units, no problem occurs. When handling a long compressed image, a wasteful area is secured, and as a result, there is a problem that the entire buffer must be unnecessarily increased.

The present invention has been made in view of the above conventional problems, and an object thereof is to provide an image processing apparatus in which the order of writing image data to a buffer and the order of reading image data from the buffer are different from each other. It is to solve the problem that the next image data cannot be stored because the free areas are dispersed in the buffer.

Another object of the present invention is to eliminate the above deadlock while reducing the image data waiting for a free area as much as possible.

[0014]

[Means for Solving the Problems]

(1) In order to achieve the above object, the present invention provides a storage unit in which a plurality of image data is written, a writing unit for writing image data in the storage unit, a reading unit for reading image data from the storage unit, An output unit for outputting the image data stored in the storage unit, wherein the output unit outputs the image data in an order different from the order in which the data is written. A start address management unit that manages a start address, and an effective free area in which image data to be stored next can be stored by referring to only a part of the buffer including the start address for searching the free area The effective empty area is not found even if the search is started from the start address and a predetermined search end condition is reached. A free area search means for ending the search, and writing the next image data there when the valid free area is found, and writing the next image data when the valid free area is not found Write control means for maintaining a wait state.

For example, when searching an effective empty area for storing the next image data on the storage unit (ring buffer), the search is started from the start address designated by the write pointer, and the search is performed on the storage unit. When the effective empty area is specified over the entire area, the next image data is written even if it is considerably distant from it regardless of the position of the write pointer or the storage area of the previously written image data. However, in such writing, as a result, there is a high possibility that each empty area after the image data is read will be scattered across the entire storage unit, and the above-described deadlock may occur.

Therefore, the present invention is characterized in that each image data is written in such a manner that empty areas continuously occur as much as possible while taking the data processing efficiency into consideration.

That is, according to the present invention, when a free area is searched from the start address, if the effective free area is not found even if the search is performed until a predetermined search end condition is satisfied, the search is forcibly terminated. To do. That is, in the present invention, the search for the free area is not always performed over the entire buffer, but is ended under a fixed condition. In other words, in the case of normal image processing, based on the empirical rule that two image data sets that have a close input order will also have a relatively close output order, the image data that has a close input order as a whole storage unit The write control is performed so that the two are not separated from each other. Therefore, the search end condition is preferably set based on the write pointer indicating the start address and based on the storage area of the image data stored previously. If the effective empty area is found before the search end condition is reached, the next image data is written therein, and if not found, the writing waiting state is maintained. After that, the image data is read from the storage unit, and when the storage area is released and an empty area is generated, the search for the effective empty area is performed again. A buffer memory or a hard disk is used as the storage unit, and a ring buffer is basically used.

(2) Further, the present invention further includes compression means for variable-length compressing the image data prior to writing to the buffer, and the free area search means considers the data amount of the image data. Then, the effective free area is specified. That is, the present invention can be applied to an apparatus in which the data amount of image data is fixed, but is particularly useful in an apparatus in which writing control of variable-length compressed image data is performed.

(3) Further, the present invention includes a reference address management means for managing a reference address corresponding to the last address of the storage area of the image data previously written, wherein the empty area search means is based on the reference address. The present invention is characterized in that the search is terminated when the effective free area is not found even after the search is performed up to a predetermined number n (n ≧ 1) free area.

With the above arrangement, the storage position of the previously stored image data is used as a reference, and a predetermined number n of empty regions from that position are searched. This makes it possible to condition that the next image data is stored in a place not far away from the storage area of the previously stored image data, and the possibility of continuous empty areas is increased. Of course, if an effective empty area exists before the previously stored image data, the next image data is stored there.

The predetermined number n is preferably selected from numerical values such as 1, 2, 3 according to various conditions. For example, in the case of n = 1, the search is performed up to the first empty area after the previously stored image data (the number of skips is 0), and in the case of n = 2, 2 is counted from the previously stored image data. The search is performed up to the second empty area (the number of skips is 1). The number of skips is the number of interlaced data blocks when performing a search when one image data or a plurality of consecutive image data is defined as one data block.

(4) Further, the present invention is characterized by further comprising setting means for setting the predetermined number n. Here, it is preferable that the setting unit sets the predetermined number n based on the degree of difference between the writing order of the image data to the storage unit and the reading order of the image data from the storage unit. Alternatively, preferably, the setting means dynamically refers to information associated with each image data to dynamically set the predetermined number n.
Set.

By thus variably setting the predetermined number n (in other words, the upper limit of the skip number) according to various conditions,
The balance between the degree of waiting for writing and the degree of dispersion of free areas can be adjusted as appropriate.

[0024]

Embodiments of the present invention will be described below with reference to the drawings. In this embodiment,
A printer device as an image processing device will be described as an example.

FIG. 1 is a schematic diagram showing a configuration example of a system to which the present invention is applied. Various printers 1A-1D
Are print servers 2A to
It is connected to 2D. The print servers 2A to 2D are
The input / output control of image data and the editing of image data are performed. The print servers 2A to 2D are connected to, for example, a local area network 4 (hereinafter, referred to as LAN) on a network set up in a predetermined area, and the LAN
It is connected to the public line 5 or the like via 4 and can communicate with other input / output devices and terminal devices (including the facsimile device 6 etc.) via these nets.

The user prints to one of the print servers 2A to 2D on the net by using a graphical user interface (GUI) in the client 3 or a print command provided in an application such as word processing software. Request (print job). The print server that receives this request outputs the image data to the printer connected to the print server or to another print server while performing the appropriate image processing by checking the content of the job. As a result, the print processing desired by the user is realized.

Direct printing is used to output image data to a printer connected to the print server.
The case where image data is output to another printer via another print server is called redirect printing.

FIG. 2 shows various printers 1A shown in FIG.
1 to 1D are cross-sectional views showing an example. The printer 1 shown in FIG. 2 will be described below for reference.

The image forming unit M1 provided inside the printer 1 forms a toner image corresponding to the original image on the recording paper based on the image information transmitted from the terminal device on the LAN 4 or the public line. And this printing unit M2
And a paper feeding unit M3 that supplies recording paper to the paper. Here, the printing unit M2 forms a toner image on a recording sheet by a well-known electrophotographic method in accordance with the digital image data from the processing unit, and the toner image of the photosensitive drum M5 uniformly charged by the charging device M4. The surface is exposed by laser light from the laser exposure device M6 to form an electrostatic latent image. The laser exposure device M6 is a laser element (not shown) such as a semiconductor laser whose drive current is modulated based on image data, and directs laser light from the laser element in a direction orthogonal to the moving direction of the surface of the photosensitive drum M5. It is composed of a rotating polygon mirror M7 that periodically deflects, a reflecting mirror M8, and the like.

The electrostatic latent image on the photoconductor drum M5 is developed by the developing devices M9 and M10 to form a toner image of a desired color on the photoconductor drum M5. This toner image is transferred to the transfer portion M.
By 11, the image is transferred onto the recording paper sent along the path A from any one of the plurality of trays M3a to 3e of the paper feeding unit M3. It should be noted that the trays M3a to 3c are paper feed trays for storing papers of different sizes, the tray M3d is an intermediate tray for temporarily storing recording papers for double-sided copying, and the M3e is for several hundred sheets. It is a large-capacity tray that stores the recording paper of. After the transfer, the residual toner remaining on the surface of the photosensitive drum M5 is removed by the cleaning unit M12.

The recording sheet after transfer is peeled from the photosensitive drum M5 by the peeling section M13, conveyed to the fixing section M15 by the conveyor M14, and subjected to the fixing process. The path of the sheet of paper after fixing is switched by the switching gate M16 to either the path B which goes to the discharging section M17 or the path C which goes to the intermediate tray M3d through the reversing section M18 for double-sided copying. In the case of double-sided copying, the front and back sides of the recording paper are reversed at the reversing unit M18 and are supplied again to the printing unit M2 along the path A via the intermediate tray M3d, and a toner image is formed on the back surface of the recording paper this time. After that, it is sent to the discharge unit M17.

FIG. 3 is a diagram showing an embodiment of the image processing apparatus according to the present invention, and is specifically a hardware block diagram in the print server 2. The print server 2 includes an image forming unit (Image Output Terminal: IOT) 1
10 (corresponding to the printer 1 shown in FIG. 1) is connected, and a network interface (Ether I / F) 101
The LAN 4 is connected via. In the present embodiment, Ethernet is shown as an example of LAN.

The print server 2 includes the above-mentioned network interface 101, a compressor (Compressor) 102 for compressing image data, and a D for transferring image data.
MA (Direct Memory Access) controller 103, CPU 104 for performing software processing, RAM / R
CPU peripheral chips such as OM / clock generator (Peri
pheral) 105, a page buffer (ring buffer) 106 for storing image data, a hard disk 108 (HD) for temporarily storing image data,
It is composed of a disk controller (DiskCont) 107 for controlling the hard disk, a decompressor 109 for decompressing image data, an internal bus 111 for transferring image data and control data, and the like.

Next, the overall operation of the print server 2 will be described. Image information from the network is usually Page Description Language (PDL).
However, since the image information handled by the IOT 110 is raster data (bitmap data), it cannot be output to the IOT 110 as it is. Therefore, when image information is input from the LAN 4, it is necessary to first convert the image information into raster data. This processing is called decomposing, and the module (software or hardware) that performs the decomposing processing is called a decomposer. The image data that has undergone the decomposing process and has been converted into raster data is compressed by the compressor 102 and temporarily stored in the page buffer 106. Since the page buffer 106 has a small capacity, the image data is transferred to the hard disk 108 and temporarily saved.

On the other hand, the compressed image data is I
When outputting to the OT 110, the image data is transferred to the decompressor 109 through writing / reading to the page buffer 106 as described later in detail, and the decompressed image data is output to the IOT 110. IOT
In 110, in the image forming unit M1 shown in FIG. 2, halftone dots are formed by controlling on / off of laser light for each pixel based on the binary data generated from the input image data. The halftone image is reproduced, and the print image is output.

The image data is output by such a series of processes. The process of decomposing the image information, the process of storing the compressed image data in the page buffer 106 while performing the compression process, the temporary saving to the hard disk 108 or the hard disk. To page buffer 106
The process of returning the image data to the IOT 110 and the process of outputting the image data from the page buffer 106 to the IOT 110 while decompressing the image data are processed in parallel by the bus arbitration and the time division process of the internal bus 111 as much as possible. As a result, the processing speed can be improved as a result.

FIG. 4 is a functional block diagram showing the control unit 10 of the print server 2. Also,
FIG. 5 shows a management table used for controlling the control unit 10.

From a hardware standpoint, the control unit 10 of FIG. 4 is, for example, C as shown by reference numeral 10 in FIG.
It is realized by the PU 104, the CPU peripheral chip group 105, and the like. Incidentally, the ROM included in the CPU peripheral chip group 105 stores a control program or the like for causing the CPU 104 to function as a control unit, and the RAM included in the CPU peripheral chip group 105 is formed with a job management table described later. It For convenience of description, FIG. 4 also shows configurations corresponding to the compressor 102, the disk controller 107, and the like.

In FIG. 4, the control unit 10 includes a job receiving unit 201 and a job scheduling unit 203,
Among them, the job queuing unit 202 acts as an intermediary for exchange for job processing.

First, the initialization unit 200 starts up the system, and the job can be accepted. At this time, the skip count selection unit 211 sets a predetermined skip count (upper limit) as a default value. The number of skips relates to a search end condition when searching for an empty area, and corresponds to the number of data block jumps when performing a search beyond a skip pointer described later. It is desirable that the number of skips be appropriately set based on the degree of discrepancy between the data input order to the page buffer and the data output order from the page buffer. For example, if the order difference is about 1 to 3 plates, the number of skips may be set to 1. If it is within 10 plates, the number of skips may be set to 2, and if it is more than that, the number of skips may be set to 3. set. In this way, the number of skips is selected according to the degree of discrepancy.If the number of skips is set to a small value even though the degree of discrepancy is large, the data write waiting frequency increases, and as a result, the processing speed increases. This is because it will decrease. When the number of empty areas counted from the skip pointer is n (n ≧ 1), the number of skips is n−1.

After the above initialization, when the job accepting unit 201 accepts a job processing request from the client via the net, the job content is analyzed there, and the analysis result is stored in the job ticket table 14 by the job content storing unit 210. (See FIG. 5). As a result, the job queuing unit 202 performs queuing, while the processing of the job receiving unit 201 waits for the next job.

The job scheduling unit 203 constantly monitors whether or not a job exists in the job queue 12 (see FIG. 5) of the job queuing unit 202.
Take the job from and check its contents. In the present embodiment, when it is determined that the job can be processed as a result of the check, the skip count selection unit 211 resets the skip count according to the job content. Thus, in this embodiment, the IO as described above is used.
The most probable skip count can be determined by optimizing the skip count depending on T, that is, the environment and the job content.

As shown in FIG. 5, the number of skips is also written in the job ticket table 14 as a part of the job content. This skip count information is
When the job is created on the client side, it can be attached to the image data as one of the information constituting the job. For example, on the print server side based on general job information such as the number of copies, double-sided / single-sided, etc. The number of skips may be calculated automatically with.

For the first time, the image generation / compression / writing unit 213 performs the above-described processing for outputting image data. At that time, the information necessary for managing the image data in the page buffer, such as an empty area and the stored image data number (Plate N
o. ), The address to write next (address pointer),
The skip pointer and the like, which will be described later, are stored in the page buffer management unit 2
14 collectively manage. Specifically, the data structure in the buffer is managed by the buffer table (BufferTable) 16 and the buffering information table (Buffering Info) 18 in the management table shown in FIG. The latter table 18 stores information about each plate stored in the buffer.

The job scheduling unit 203 gives a command to the image reading unit 216 via the page buffer management unit 214 when preparation for output is completed. As a result, after the image data is written from the hard disk to the page buffer in a predetermined order, the output processing control unit 217 reads the image data from the page buffer in a predetermined order and performs an output process.

More specifically, in order to secure a storage area for each image data read from the hard disk, the job scheduling unit 203 requests the page buffer management unit 214 to search for a free area and specify an effective free area. To do. The search for the empty area is performed by the skip processing (free area search) unit 215, and if the skip processing is required at the time of searching for the empty area, the skip count set in the skip count selection unit 211 is set as the upper limit. As such, the skip processing is performed. Here, if an effective empty area in which the next image data can be stored is found before the actual skip count reaches the set skip count, the write control of the image reading unit 216 causes
The next image data is written in the empty area. On the other hand, when such an effective free area is not found, the write waiting state is maintained. The connections between these modules will be clarified in the flowcharts shown in FIGS. As is clear from the above description, the page buffer management unit 214 and the image reading unit 216 function as a writing control unit, and the output processing control unit 21.
Reference numeral 7 functions as a reading means.

Next, the page buffer management according to the present invention will be specifically described with reference to FIGS. 6 and 7.

A buffer management model is shown in FIG. The page buffer 106 is a so-called ring buffer, and its start address 24 to end address 26
Up to is the entire buffer area 22. When the head address 24 is indicated by an absolute address, it indicates a base address, and when it is indicated by a relative address, it is an address 0. Similarly, when the final address is indicated by an absolute address, it indicates an address obtained by adding the buffer entire region to the base address, and when indicated by an absolute address, it indicates an address obtained by subtracting 1 from the buffer entire region. The buffer management unit 214 shown in FIG.
Basically uses relative addresses.

Reference numeral 18 indicates a storage area for each image data (plate). The start address and size of the storage area of each plate are stored in the buffering information table 18 shown in FIG. In relation to this, as information for the skip processing (empty area search) unit 215 to search for an empty area, is it read out for each plate in the buffering information table 18 and can be overwritten (ON), Alternatively, information that has not been read yet and cannot be overwritten (OFF) is also stored.

The write pointer P indicates the start address when searching for an empty area for writing the next image data (plate). The write pointer P shifts to the first address of the next empty area when the image data is written from the currently indicated address. For example, in the state where the write pointer P exists as shown in FIG. 6, when another plate is written after the plate 2, the write pointer P moves to the first address of the next empty area. However, in the present embodiment, when the image data is written by the skip processing, the write pointer does not move and is maintained at the original position.

Although the final address and the start address are regarded as consecutive addresses in the search, the image data cannot be stored over the end address and the start address for the convenience of DMA or the like. When the write pointer P and the final address 26 approach each other up to a predetermined address length, the write pointer P wraps around from the final address to the start address side and jumps to the start address of the first free area.

The skip pointer S indicates the final address of the plate stored by the previous skip processing. When the plate is written by the skip processing, the skip pointer S moves to the final address of the plate.

FIG. 7 shows an example of writing the plate in the buffer management model as described above. In this example, plates 1, 2, 3, 4, 5, 6,
Writing is performed in the order of 7, 8 ...
Reading is performed in the order of 6, 1, 3, 5, ....
Further, in this example, 1 is set as the upper limit of the skip count.

First, in (A), the page buffer 106
The plates 1, 2, 3, 4, 5, 6, 7 are stored in.
There is a relatively small free space 30 after the plate 7, but not enough to accommodate the plate 8. Therefore, the plate 8 is placed in a write waiting state. The write pointer P indicates the first address of the free area 30.

From this state, as shown in (B), when the plate 2 is read and the area 32 is released, an area sufficient for storing the plate 8 is triggered by the change in the buffer content. A search for an effective free area is carried out. In this case, the empty area 30 indicated by the write pointer P cannot still write on the plate 8 due to capacity, but the empty area 32 existing at the position where the plate 1 is skipped (the original storage area of the plate 2) It is possible to write a plate 8 on it. Therefore, the plate 8 is written in the area 32 in a right-justified manner. Note that this creates a small empty area 34 between the plate 8 and the plate 3. Here, the skip pointer S is set to the final address of the plate 8 written in the skip processing.

If the empty area 32 does not have a sufficient capacity to store the plate 8, the skip number has already reached its upper limit (= 1), so that there is an effective empty area even after that. If so, the search ends there, and the plate 8 continues to be in a write wait state. This is because of the continuous free area as described above.

From the state of (B), as shown in (C),
When the storage area 36 is released by reading the plate 4, a search for an effective free area for storing the plate 9 is executed. In this case, it is determined that the empty area 30 cannot be stored in terms of capacity, and the empty area 34 found by skipping the data block of the plate 1 and the plate 8 connected to it is also not possible in terms of capacity. However, since the plate 9 can be stored in the empty area 36 (original storage area of the plate 4) that exists where the plate 9 is subsequently skipped, the plate 9 is written in the empty area 36 in a left-justified manner. Be done.

Here, the skip number is counted as the skip number after the skip pointer S, that is, the skip number before the skip pointer S is irrelevant.

(D) shows the state when the buffer management is advanced according to the above rules. In this state, the write pointer P exists at the beginning of the empty area 38, and the skip pointer S exists at the final address of the plate 17. Further, after the plate 17, empty areas 40, 42, 4 formed as an effect of the skip process are formed.
There are 4 continuums.

In this state (D), if an attempt is made to store a plate 18 having a relatively large amount of data, it is impossible to store it in the first found free area 38, but the plates 15, 16 and 17 cannot be stored. The empty area 40 found after skipping the data block is the next empty area 42.
, And the plate 18 can be stored sufficiently. Therefore, the plates 18 are stored in the two areas in the left-justified manner.

FIG. 8 shows, as a comparative example, the result of performing the write control of each plate without defining the search termination condition. In this method, as writing and reading of the plate gradually progresses, the previously stored ones and the later stored ones are mixed, and the released ones are further released. 46,
As shown by 48, 50 and 52, there is a high possibility that the free areas will be dispersed. As a result, even if an attempt is made to store the plate 18 and an effective free area is searched from the write pointer P (there is no upper limit of the number of skips), although there are many free areas as a whole buffer, in the end,
The plate 18 cannot be written in any of the empty areas. That is, deadlock has occurred in this comparative example.

On the other hand, according to the present embodiment, when the empty area is searched from the write pointer P, the upper limit of the number of skips (the upper limit of the number of empty areas to be searched) is set based on the skip pointer S. , Two plates having a close writing order can be stored in the page buffer without being separated from each other, and as a result, the probability of continuous empty areas can be increased. This is based on the empirical rule that the read order will be close between the plates whose write order is relatively close, and deadlock can be effectively prevented even though the write control is simple.

Of course, it is desirable to set the upper limit of the number of skips as appropriate so that a write wait does not occur and the data processing speed does not decrease so much.

Next, the operation of the image processing apparatus of this embodiment will be described with reference to FIGS. FIG. 9 is a flowchart showing the operation of the job receiving unit 201 shown in FIG. First, in S1, it is determined whether or not there is a job processing request from the client. If there is a job processing request, the job content is confirmed in S2, and if processing is possible, the job content is stored in the job management table (job ticket 14). To do. In S3, a default value is set as the upper limit of the skip count. This value is appropriately selected depending on the degree to which the principle of releasing in the order of storage can be inconsistent with the principle. And S
In step 4, the job ticket 14 is queued, in step S5, the job scheduling unit 203 is notified that there is a job request, and the process returns to step S1 to wait for the next job request.

FIG. 10 is a flowchart showing the operation of the job scheduling unit 203 shown in FIG.

In step S6, when the notification is received from the job receiving unit 201, the job ticket 14 is transferred from the job queue 12.
Take out. Then, in S7, the contents of the job are checked to dynamically check whether or not an appropriate number of skips is given. In S8, the image data is processed so that it can be output. Details of this process are shown in FIG. In S9, in order to output the image data, the image data is written in the page buffer. The details of the processing in this case are shown in FIG. In S10,
It is checked whether the above S8 and S9 are ready and output is possible. If not, the process proceeds to S11,
If possible, S12 is executed. In S11, it is checked which of the preparations in S8 and S9 is not completed, and S8
Alternatively, the process is returned to S9. Then, in S12, the output process of the image data is executed. In S13, it is checked whether or not the output of all the image data is completed, and if not, the process returns to S10, and if all the output is completed, S1
Proceed to 4. In this S14, the image data in the hard disk is erased. In S15, the job ending process is performed, and specifically, the job ticket is deleted.

Note that S16 represents a module that performs output processing by starting the output processing in S12.
It operates in parallel with the main processing (operations shown in FIGS. 9 to 12). The content of the output process is as described based on FIGS. 2 and 3.

FIG. 11 shows the operation in the image processing section corresponding to S8 shown in FIG.

First, the decomposer is activated in S21, and the decompose process is executed in S22. This processing is similar to the output processing of S16, and is the main processing (see FIGS.
2) can be operated in parallel. Since the image information sent from the client is in the page description language, raster data is generated here by decomposing. In S23, an area for storing the decomposing result is requested on the page buffer. In this case, in the determination of S24, if the area can be secured, the process proceeds to S26, and if not, the process proceeds to S25. Note that this area search is performed prior to compression of image data and writing to the hard disk, that is, because the release is performed in the same order as the order in which the image data was stored,
The above skip processing is not performed. In S25, a standby state is waited, and in S26, the process waits for raster data to be stored in the page buffer in S22. Then, in S27, the stored raster data is compressed.
In S28, the process of writing the compressed image data that has been compressed to the hard disk is started. Note that the processing of writing to the hard disk in S29 is the same as the main processing (FIGS. 9 to 12) as in the decomposing processing of S22 and the output processing of S16.
Operation) is executed in parallel.

At S30, when the process of writing to the hard disk is completed, the buffer can be overwritten, so the buffer management table is updated. On the other hand, in S31, when waiting for the buffer to be free in S25, it is checked in S24 whether or not the buffer can be secured again based on the update result. In S32, it is checked whether the writing process has been completed for all the plates, and if not, S
Returning to step 21, the process is repeated, and when the process is completed, this flow ends.

FIG. 12 shows the operation of the image processing section corresponding to S9 shown in FIG.

In S41, the process of reading the compressed image data from the hard disk is started. Here, the processing of reading from the hard disk shown in S42 is the same as the writing processing of S29 (main processing (operation shown in FIGS. 9 to 12)).
Works in parallel with. In S43, an area for storing the read result is requested on the page buffer. In this S43, immediately after the write pointer indicating the start of search,
If there is not a sufficient area for writing the image data, the skip processing is subsequently performed while maximizing the number of skips set in the buffer management table. The skip processing is as described based on FIGS. 5 and 6. If the area can be secured in S44, the process proceeds to S46, and if not, the process proceeds to S45.
Proceed to. In S45, if the effective free area cannot be secured even after the skip processing is performed, the free area wait state is entered.
In S46, the process waits for the compressed image data to be read onto the buffer in S42.

S47 is a module for performing the output process shown in S16, and every time the output of one page is completed, the fact is notified. In S48, the buffer management table is updated based on the information indicating the end of output of one page. In S49, when waiting for the buffer to be free in S45, it is determined whether or not the buffer can be secured again based on the update result.
Check at 4. Then, in S50, it is checked whether or not all the outputs are completed, and if not, S4.
When the process returns to 1 and the process is repeated, and the process ends, this flow ends.

Note that, in actual input / output control of image data, error cancellation processing, communication processing, and the like exist in addition to the above, but they are omitted because they are not related to the present invention. Further, in the above embodiment, the printer is taken as an example, but if the image processing apparatus has a buffer,
The present invention can be applied to any image processing device such as a copying machine, a printer device, a facsimile device, and the like.

[0075]

As described above, according to the present invention,
In an image processing device in which the order of writing image data to the buffer and the order of reading image data from the buffer are different, it is possible to solve the problem that the empty area is dispersed in the buffer and the next image data cannot be stored. Also, deadlock can be effectively prevented while minimizing the amount of image data waiting for a free space.

[Brief description of drawings]

FIG. 1 is a schematic diagram showing a system configuration to which the present invention is applied.

FIG. 2 is a sectional view showing a schematic configuration of a printer to which the present invention is applied.

FIG. 3 is a hardware block diagram showing an embodiment of an image processing apparatus according to the present invention.

FIG. 4 is a diagram showing a configuration of a control unit according to the present invention.

FIG. 5 is a diagram showing a management table.

FIG. 6 is a diagram showing a buffer management model.

FIG. 7 is a diagram showing buffer write control according to the present invention.

FIG. 8 is a diagram showing a comparative example when skip processing is not performed.

FIG. 9 is a flowchart illustrating a control procedure according to the present invention.

FIG. 10 is a flowchart illustrating a control procedure according to the present invention.

FIG. 11 is a flowchart illustrating a control procedure according to the present invention.

FIG. 12 is a flowchart illustrating a control procedure according to the present invention.

[Explanation of symbols]

1A-1D printer, 2A-2D print server,
102 compressor, 106 page buffer, 109 decompressor, 201 job receiving unit, 202 job queuing unit, 203 job scheduling unit, 211
Skip count selection unit, 214 page buffer management unit,
215 skip processing unit, P write pointer, S
Skip pointer.

Claims (6)

[Claims] 1. A storage unit in which a plurality of image data is written, a writing unit for writing image data in the storage unit, a reading unit for reading image data from the storage unit, and an image data stored in the storage unit. In the image processing apparatus, which includes an output unit for outputting, and outputs in an order different from the written order, the writing unit includes a start address management unit that manages a start address when searching a free area on the storage unit. A means for identifying an effective empty area in which image data to be stored next can be stored by referring to only a part of the buffer including the start address for searching the empty area, Empty to start the search and end the search if the effective free space is not found even if the search is performed until a predetermined search end condition is reached Area search means for writing the next image data therein when the valid free area is found, and for maintaining the waiting state for writing the next image data when the valid free area is not found An image processing apparatus comprising: a control unit. 2. The apparatus according to claim 1, further comprising a compression unit that compresses image data in a variable length prior to writing to the storage unit, and the empty area search unit considers a data amount of the image data. The image processing apparatus is characterized by specifying the effective free area. 3. The apparatus according to claim 1, further comprising a reference address management unit that manages a reference address corresponding to a final address of a storage area of the image data written last time, wherein the empty area search unit includes the reference address. An image processing apparatus characterized by terminating the search when the effective empty area is not found even when the search is performed up to a predetermined number n (n ≧ 1) empty area counted from. 4. The image processing apparatus according to claim 3, further comprising setting means for setting the predetermined number n. 5. The apparatus according to claim 4, wherein the setting unit sets the predetermined number n based on the degree of difference between the order of writing image data to the buffer and the order of reading image data from the buffer. An image processing device characterized by: 6. The image processing apparatus according to claim 4, wherein the setting unit dynamically sets the predetermined number n by referring to information associated with each image data.