JP2016009268A - Image processing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the overhead of an overlapping determination time while accelerating a drawing speed by omitting an overlapping area to execute drawing flattening processing when overlapping exists in an object.SOLUTION: An image processor includes flattening processing means 26, 28 for performing flattening processing of image data with the number A (A is an integer of two or more) of pieces of image data as a group, and rasterization processing means 32 for generating a raster image from the image subjected to the flattening processing. Processing of image data constituting a group with a group number as N (N is an integer of one or more) in the rasterization processing means 32, and processing of image data constituting a group N+1 in the flattening processing means 26, 28 are executed in parallel.

Description

  The present invention relates to an image processing apparatus.

  Patent Document 1 describes a drawing processing device that performs an efficient object drawing process by checking an overlapping area of intermediate data (objects) in advance. The intermediate data distribution unit determines whether there is an overlapping area between the intermediate data and the unexpanded intermediate data, and distributes the intermediate data having no overlapping area to the expansion processing unit based on the priority order.

  Japanese Patent Application Laid-Open No. 2004-228561 describes that before the object overlap determination process, the determination target is thinned out by classifying whether to flatten or not according to the size of the bounding box.

JP 10-333852 A JP 2010-157040 A

A flattening process that determines whether or not there are overlapping objects and performs drawing while omitting the overlapping area is effective in making the drawing process more efficient. There is a problem that takes time. That is, if there are N objects, the presence / absence of overlap is analyzed in all ways, so the number of analyzes is _N C ₂ = N (N−1) / 2, and the overlap determination time per time is α. Then, a total time of α · N (N−1) / 2 is required, and the overlap determination time can be an overhead.

  When the presence / absence of flattening is classified according to the size of the bounding box before the object overlap determination process, the number of times of analysis is reduced as compared with the whole method, but the overlap determination time may remain as overhead as well.

  An object of the present invention is to provide an apparatus capable of suppressing the overlap determination time overhead while increasing the drawing speed by performing a flattening process for drawing without overlapping areas when objects overlap. .

  According to the first aspect of the present invention, flattening processing means for performing flattening processing on the image data by grouping the number A of image data (A is an integer of 2 or more), and a raster image from the flattened image data. And a rasterizer processing means for generating, the group number is N (N is an integer of 1 or more), the processing of image data constituting the group N in the rasterizer processing means, and the group N + 1 in the flattening processing means An image processing apparatus that executes image data processing in parallel.

  According to the second aspect of the present invention, a flattening processing means for performing a flattening process on the image data by grouping the number A of image data (A is an integer of 1 or more), and a raster image from the flattened image data. And a rasterizer processing means for generating, the group number is N (N is an integer of 1 or more), the processing of image data constituting the group N in the rasterizer processing means, and the group N + 1 in the flattening processing means Prior to processing of the image data constituting the group N in the rasterizer processing means, the group N + 1 in the flattening processing means is configured according to the expected processing time. Processing time of image data in the rasterizer processing means Such that less processing time of the image data constituting the loop N, variably setting the image data number A constituting the group N + 1, an image processing apparatus.

  According to the third aspect of the present invention, flattening is performed when the expected rendering time of the image data is relatively long, and flattening is not performed when the image data is relatively short. Flattening processing means for performing flattening processing and rasterizer processing means for generating a raster image from flattened image data and non-flattened image data, and image data in the rasterizer processing means And an image processing apparatus that executes the processing of the image data in the flattening processing means in parallel.

  According to the first to third aspects of the present invention, when the object overlaps, the flattening process for performing drawing while omitting the overlapping area is performed to increase the drawing speed and suppress the overhead of the overlap determination time. it can.

It is a schematic diagram of a flattening process. It is analysis / drawing process explanatory drawing of a conventional system. It is an analysis and drawing process explanatory drawing of embodiment. It is a block diagram of the configuration of the first embodiment. It is a processing flowchart of a 1st embodiment. It is a graph which shows the processing time of a flattening process and a rasterizer. It is a schematic diagram which shows the relationship between a flattening process and a rasterizer. It is a schematic diagram which shows the relationship between a flattening process and a rasterizer. It is a schematic diagram which shows the relationship between a flattening process and a rasterizer. It is a block diagram of the configuration of the second embodiment. It is a block diagram of the configuration of the second embodiment. It is a graph which shows the relationship between an object and drawing time. It is a graph which shows the relationship between an object and drawing time. It is a conceptual block diagram of 3rd Embodiment. It is a block diagram of the configuration of the third embodiment. It is a processing flowchart of a 3rd embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

<First Embodiment>
First, the flattening process will be described prior to the description of the embodiment. The print data described in the page description language PDL is converted into intermediate data to enable high-speed expansion processing. If there is an overlap between objects that are the minimum unit of the intermediate data, the newly drawn object is already drawn. Since the overwritten object is overwritten, the overlapping portion is drawn redundantly, which is a waste of processing. Therefore, when there is an overlap between objects, the drawing speed can be increased by omitting the drawing of the overlapping portion. This is called flattening processing.

  FIG. 1 is a schematic diagram of the flattening process. When there is an overlap between the object 10 and the object 12 expressed by a rectangle, a drawing pattern is created by omitting (or removing) the overlap portion from the object 10 so as to be drawn first in time series. Draw. Since the overlapping portion is only drawn as the object 12, the drawing speed is increased as compared with the case of overlapping drawing.

However, in order to determine whether or not there is an overlap between all objects to be printed, if the total number of objects is N, the total number of combinations
_N C ₂ = N (N−1) / 2
Analysis is required.

FIG. 2 is a schematic diagram for determining whether or not there is an overlap between objects with respect to the total number N of objects. As shown in FIG. 2 (a), if an object of a processing unit is composed of objects 1 to N, _N C ₂ = N (N−1) / 2 in order to analyze these N objects. Analysis is required, after which N objects are drawn. Therefore, as shown in FIG. 2B, the total processing time is the sum of the time required for the flattening process and the time required for the drawing, and the time required for the flattening process, that is, the analysis time is exposed as it is. Become. This analysis time increases as the number N of objects increases.

  Therefore, in the present embodiment, the number of analyzes is reduced by grouping a certain number of objects, not all, and performing flattening processing within the group.

FIG. 3 is a schematic diagram of flattening processing and drawing processing in the present embodiment. A plurality of objects are grouped together and a flattening process is performed using these groups as processing units. For example, as shown in FIG. 3A, three objects are grouped together.
Object 1, Object 2, Object 3 → Group 1
Object 4, Object 5, Object 6 → Group 2
Object 7, object 8, object 9 → group 3
...
It is like this. Then, it is determined whether or not there is an overlap in the group, and the flattening process of the next group is performed in parallel while the flattening process of a certain group is completed and the drawing process is executed.

  FIG. 3B is a schematic diagram of flattening processing and drawing processing for each group executed in parallel. In the figure, the number indicates the number of a group consisting of three objects. A flattening process including an overlap analysis of group 1 is performed, a drawing process of group 1 is performed, and a flattening process of the next group 2 is performed concurrently with the drawing process. The flattening process for group 2 is completed, the drawing process for group 2 is performed, and the flattening process for the next group 3 is performed in parallel with this drawing process. Thereafter, the flattening process of the group N-1 is performed in the same manner, the drawing process of the group N-1 is performed, and the flattening process of the next group N is performed in parallel with this. By limiting the flattening process of group N to a certain number of objects, the time required for the flattening process is suppressed, and the flattening process of group N is performed in parallel with the drawing process of group N-1. The flattening process of group N can be concealed by the drawing process of group N-1, and the overhead can be suppressed by concealing the flattening process.

  In the present embodiment, a group of a certain number of objects, for example, three objects, can be said to be an object window that defines a processing unit for flattening processing and drawing processing.

  FIG. 4 is a block diagram showing the configuration of the image processing apparatus according to this embodiment. The image processing apparatus includes an input control unit 20, an object analysis unit 22, a processing number determination unit 24, an overlap determination unit 26, an overlap processing unit 28, a processing window control unit 30, a rasterizer drawing unit 32, and an output. A control unit 34 and a reconstruction control unit 36 are provided.

  The input control unit 20 inputs intermediate data (object) supplied from the outside and outputs it to the object analysis unit 22. An object is represented by a set of simple figures such as rectangles. In the case of a rectangle, the arrangement of the object is managed by the coordinates of the corners of the rectangle. For example, it is managed as section data representing a section that an object occupies on one scanning line. Each section data includes the coordinates of the start point and end point of the section, and the object type (character, graphic, Information indicating a continuous tone image), pixel value attributes on section data, and the like.

  The object analysis unit 22 analyzes the section data of the object and outputs it to the processing number determination unit 24. The object analysis unit 22 calculates the time required to draw the object from the section data of the object as necessary.

  The processing number determination unit 24 determines the number of objects to be flattened. This number is a preset number and is an integer of 2 or more. For example, the processing number determination unit 24 sets the number of objects to three. The number of objects may be stored in advance in the memory of the image processing apparatus, or may be input by an operator of the image processing apparatus from an input device such as a keyboard.

  The overlap determination unit 26 determines whether there is an overlap between the three objects based on the section data. Specifically, it is determined whether or not the rectangular coordinates between the objects overlap each other. Assuming that the three objects are object 1, object 2, and object 3, the overlap determining unit 26 determines the overlap between object 1 and object 2, the overlap between object 2 and object 3, and the overlap between object 1 and object 3. The overlap determination unit 26 outputs the determination result to the overlap processing unit 28.

  When the objects overlap each other, the overlap processing unit 28 generates a mask for the overlapping portion and outputs the mask to the rasterizer drawing unit 32. For example, when the object 1 and the object 2 overlap, a mask is generated for the overlapping portion of the object 1. If the objects do not overlap, do not generate a mask. In addition, the overlap processing unit 28 outputs the processing result to the processing window control unit 30.

  The processing window control unit 30 controls the processing window according to the processing result in the overlap processing unit 28. Specifically, when the overlap processing of the group 1 composed of three objects is completed, the three objects constituting the next group 2 are designated and output to the overlap determination unit 26. The three objects constituting the group 2 are the object 4, the object 5, and the object 6, and the overlap determining unit 26 determines the overlap between these objects. Hereinafter, similarly, the processing window control unit 30 sequentially designates the three objects constituting the groups 3, 4,.

  The rasterizer drawing unit 32 generates a raster image from the flattened object and outputs the raster image to the output control unit 34. When there is no overlap, the rasterizer drawing unit 32 generates a raster image as it is from the object as in the conventional case, but when there is an overlap and a mask is generated, generation of the raster image in the mask portion is omitted. (See FIG. 1). The output control unit 34 performs printing by outputting a raster image to a print engine (not shown).

  The flattening process is executed by the overlap determination unit 26 and the overlap processing unit 28, and the drawing process is executed by the rasterizer drawing unit 32. At this time, the drawing process of the group N-1 is executed by the rasterizer drawing unit 32. In parallel with this, the flattening process of the group N is executed. When the flattening process for group N is completed, the rasterizer drawing unit 32 subsequently executes the drawing process for group N. In parallel with this, the flattening process for group N + 1 is executed.

  The reconfiguration control unit 36 reconfigures a circuit to be used in accordance with the drawing format when there are a plurality of drawing formats of the object. A DRP (Dynamic Reconfigurable Processor) capable of dynamically reconfiguring an internal logic circuit configuration is known, and a plurality of configuration data stored in a memory is selectively validated, and the DRP is validated according to the validated configuration data. By reconfiguring, it is possible to adapt to a plurality of drawing formats without increasing the hardware. If the drawing format is fixed to one type, it is not necessary to reconfigure, so the reconfiguration control unit 36 is unnecessary. When the rasterizer rendering unit 32 sequentially performs object rasterizer processing, color matching processing, and data compression processing, these series of pipeline processing are configured by DRP. Each part of FIG. 4 is configured by a computer, but DRP is configured as an auxiliary processing device of the computer.

FIG. 5 is a processing flowchart of the present embodiment. In the processing flowchart, an object group is described as an object window.
First, the window size A (2 ≦ A) in the processing number determination unit 24 is set (S101). Although the time required for the flattening process varies depending on the window size A, details will be described later.

  After the window size A is set, the overlap processing (flattening processing) of the Nth window and the rasterizer processing (drawing processing) of the (N-1) th window are executed in parallel (S102a, S102b). That is, the drawing process for the first window and the flattening process for the second window are executed in parallel, and the drawing process for the second window and the flattening process for the third window are executed in parallel. . When the flattening process of the first window is performed, the drawing process is not performed in parallel because there is no data to be drawn yet. The above process is repeated until all the objects are completed (S103). The processing window control unit 30 determines whether or not all objects have been completed.

  As described above, in the present embodiment, the number of objects to be flattened is fixed to A, and the flattening process and the drawing process are executed in parallel to conceal the flattening process in the drawing process. The overhead of flattening processing can be eliminated and the total processing time until drawing can be suppressed.

Second Embodiment
In the first embodiment, since the flattening process is executed with the window size A fixed, the analysis time for whether or not there is an overlap is constant. On the other hand, since the size is different for each object and the drawing method can be different, the drawing time can be changed for each object. Therefore, depending on the object, even if the flattening process and the drawing process are executed in parallel, the flattening process may not be concealed as a result because the drawing time is short.

  FIG. 6 is a diagram illustrating an example of the relationship between the window number and the processing time. When window size A = 16, window number 1 is composed of object 1, object 2, object 3,..., Object 16, and window number 2 is composed of object 17, object 18,. Become.

  In FIG. 6, the symbol a indicates the flattening processing time of each window, and the symbol b indicates the drawing processing time of each window. Since the window size is constant, the flattening processing time is constant regardless of the window number. On the other hand, the drawing processing time varies depending on the window number. In particular, the processing time for window numbers 3 to 6 and 14 to 17 is used. It is relatively long, and the processing time is relatively short in the window numbers 7 to 13. When the flattening process and the drawing process are executed in parallel, if the flattening process time is equal to or less than the drawing process time, the flattening process can be hidden by the drawing process. If so, the flattening process cannot be concealed.

  FIG. 7 is a schematic diagram when the flattening process and the drawing process are executed in parallel. Focusing on window number 6 and window number 7, the drawing process of window number 6 and the flattening process of window number 7 are executed in parallel, and the flattening process is concealed because the flattening process time ≤ the drawing process time. Can do.

  On the other hand, paying attention to window number 7 and window number 8, the drawing process of window number 7 and the flattening process of window number 8 are executed in parallel, and the flattening process is concealed because flattening process time> drawing process time. Can not.

  If attention is paid to window number 8 and window number 9, the drawing process of window number 8 and the flattening process of window number 9 are executed in parallel, and since the flattening process time> the drawing process time, the flattening process is similarly performed. Cannot be concealed.

  Thus, in view of the fact that the drawing time may change for each object due to the fact that the drawing time may change for each object, and the flattening processing time cannot be concealed when the flattening processing time> the drawing processing time. In the embodiment, the flattening processing time is adaptively changed by making the window size A variable (that is, variable number of objects) instead of being fixed.

FIG. 8 is a schematic diagram of flattening processing and drawing processing by the object number variable method according to the present embodiment. Now, as shown in Fig. 8 (a),
(Flatening processing time of window N) <(drawing time of window N-1)
(Flattening processing time of window N + 1)> (drawing time of window N)
(Flatening processing time for window N + 2)> (drawing time for window N + 1)
Suppose that At this time, as shown in FIG. 8B, for the flattening process of the window N, the window size A is increased so as to be approximately equal to the drawing time of the window N−1, and for the flattening process of the window N + 1, the window N The window size A is reduced so as to be substantially equal to the drawing time of the window N, and the window size A is reduced so as to be substantially equal to the drawing time of the window N + 1 for the flattening processing of the window N + 2. In this way, by adjusting the window size A for performing the flattening process, that is, the number of objects, to increase or decrease the flattening process time, the flattening process is concealed according to the drawing time. Since the size and attributes of a plurality of objects constituting the window N are analyzed during the flattening process of the window N, the estimated drawing time of the window N can be calculated at the same time as the flattening process of the window N. The window size A of the window N + 1 is variably set according to the expected time. That is, the window size A of the next window N + 1 is set during the flattening process of the window N.

FIG. 9 is a schematic diagram for setting the window size A of the window N + 1. As shown in FIG. 9A, when the flattening process of the window N is executed, the expected drawing time TN of the window _N is calculated (at this time, the drawing of the window N is not performed). Therefore, the window size A of the window N + 1, is set to be within the drawing expected time T _N of the window N. Assuming that the flattening processing time after variably adjusting the window size A of the window N + 1 is t _{N + 1} , T _N ≧ t _{N + 1} and the flattening processing of the window N + 1 is concealed.

Specifically, if the window size of the window N + 1 is A, that is, the number of objects is A, and the time required for one flattening process is α,
t _{N + 1} = α · _A C ₂ = α · A (A-1) / 2
Because
T _N ≧ t _{N + 1}
To meet
A ≦ {1+ (1 + 8T _N / α) ^1/2 } / 2 (1)
The window size A, that is, the number of objects A may be set so that As an example, the maximum number satisfying the inequality (1) is set as the window size A.

  FIG. 10 is a block diagram showing the configuration of the image processing apparatus according to this embodiment. The difference from FIG. 4 is that a next analysis number determination unit 29 is newly added.

Next analysis number determining unit 29, when the flattening processing window N, using the drawing expected time T _N of the window N, the window size A so as to satisfy the above inequality (1), that is, to determine the object number A It outputs to the processing number determination part 24. The estimated drawing time _TN is calculated by the object analysis unit 22.

  FIG. 11 is a processing flowchart of the present embodiment. First, the window size A (2 ≦ A) in the processing number determination unit 24 is set to the initial value A1 (S201). The initial value A1 may be arbitrary, and is set to 3 or 16, for example.

After the window size A is set to the initial value A1, the Nth window object A _N overlap processing (flattening processing) and the N−1th window object A _N−1 rasterizer processing (drawing processing) Are executed in parallel (S202a, S202b).

Then, to calculate the N-th window rasterizer processing number _{A N} (S203), determines a flattening process number _{A N + 1} of the (N + 1) th window using the process number _{A N} (S204). Since the number of flattening processes is substantially the same as the flattening process time, it can be said that S203 calculates the rasterizer processing time. In S204, A _{N + 1} is determined so as to satisfy the inequality (1). The above process is repeated until all objects are completed (S205). Accordingly, the first window is flattened and drawn with the initial value A1, while the second window is processed according to the number of flattening processes (processing time) of the first window. The number is determined (may be the same as the initial value A1 or may be different from this), and the flattening process and the drawing process are executed with the determined number. Similarly, for the third window, the number of processes is determined according to the number of flattening processes (flattening process time) of the second window.

  Thus, in this embodiment, the flattening process can be reliably concealed by variably adjusting the number of objects to be flattened according to the drawing time. Of course, if it is clear that the object attributes are almost uniform and the flattening processing time ≦ the drawing time, the number of objects can be fixed at the initial value A1 as shown in the first embodiment. Good.

As is clear from inequality (1), A is also increased as T _N is increased, because A also decreases as T _N is decreased, it can be said that there is a positive correlation between the two.

Applicant
(I) 500 objects with 30% overlap in the page (ii) 1000 objects with 50% overlap in the page (iii) 500 objects with 70% overlap in the page In each case, the total processing time of the conventional method (the method shown in FIG. 2) is compared with the total processing time of the method of this embodiment, and it is confirmed that the following results are obtained.
Case (i) Conventional method: 1.4 seconds Embodiment method: 0.18 seconds (ii) Conventional method: 5.2 seconds Embodiment method: 0.35 seconds (iii) Conventional method: 126 seconds Embodiment method: 1.7 seconds

  From these results, the effect of shortening the processing time or increasing the processing speed according to the present embodiment is clear.

<Third Embodiment>
In the second embodiment, the flattening process can be reliably hidden by variably adjusting the window size, that is, the number of objects, but the flattening process can be made more efficient.

  FIG. 12 is a diagram illustrating an example of the drawing time for each object. Since the size and drawing method can be different for each object, the drawing time is also different. For example, the objects 1 to 100 have a relatively long drawing time, and the objects 110 to 230 have a relatively short drawing time. The drawing also shows an example of the object number variable adjustment in the second embodiment. The first is 2, the next is 12, the next is 29, the next is 46, and the next is 111 (changes), but object m, object m + 1, object m + 2,. As described above, since the flattening process is performed only between a plurality of continuous objects, the effect of the flattening process is less likely to occur compared to the case where all the overlaps are considered. In order to maximize the effect of the flattening process, as shown in FIG. 13, it is efficient to perform the flattening process between the area A and the area C with a relatively long drawing time. However, even if the flattening process is performed on the relatively short region B or region D, the effect is inevitably limited.

  In view of the above facts, in this embodiment, the flattening process is performed for an area with a relatively long drawing time, and the flattening process is not performed for an area with a relatively short drawing time, thereby further reducing the total processing time. Shorten. An area having a relatively long drawing time can be extracted by calculating the drawing time of the object from the size and drawing method of the object and comparing it with a threshold value. Explaining with reference to FIG. 13, the area A and the area C are extracted, and the flattening process is executed on the object constituting the area A and the object constituting the area C.

  According to this method, flattening processing can be performed even for non-continuous objects.

  FIG. 14 is a conceptual configuration diagram of the present embodiment. The image processing apparatus according to the present embodiment includes a threshold Sth determination unit 40, a flattening processing unit 42, and rasterizer drawing units 44 and 46.

  The threshold value Sth determination unit inputs an object as intermediate data (intermediate language), and compares the estimated drawing time calculated from the size and drawing method of the object with a predetermined threshold value Sth. For an object equal to or greater than the threshold value Sth, the flattening processing unit 42 executes the flattening process as a flattening process target, and then the rasterizer drawing unit 44 executes the drawing process. On the other hand, with respect to an object less than the threshold value Sth, the rasterizer rendering unit 46 executes the rendering process without performing the flattening process as an untargeted (non-target) flattening process, and outputs the object.

  FIG. 15 is a block diagram showing the configuration of the image processing apparatus according to this embodiment. The difference from FIG. 10 is that an overlap omission target determination unit 23 is newly added.

The overlap omission target determination unit 23 corresponds to the threshold value Sth determination unit 40 in FIG. 14, compares the expected drawing time obtained by analyzing the object with the object analysis unit 22 with the threshold value Sth, and uses the result for the overlap omission target. Set a flag for the object. Of course, a flag may be set for an object that is not subject to overlap omission. In short, each object is classified as flattening target / non-target by comparing with a threshold value Sth. An object classified as a target for flattening processing with an estimated drawing time equal to or greater than the threshold value Sth is flattened by the overlap determination unit 26 and the like, and is drawn by the rasterizer drawing unit 32 as in the second embodiment. On the other hand, an object that has not been classified as a flattening process target because the estimated drawing time is less than the threshold value Sth is drawn by the rasterizer drawing unit 32 without being flattened by the overlap determination unit 26 or the like.
For example, for objects 1 to 6,
Object 1: Target Object 2: Target Object 3: Untargeted Object 4: Untargeted Object 5: Target Object 6: When classified as a target, a window is composed of object 1, object 2, object 5, and object 6 and is flat. Processing is executed. Object 1, object 2, object 5, and object 6 are not continuous but are subject to overlap analysis, and if there is an overlap, flattening processing is performed.

  FIG. 16 is a process flowchart of the present embodiment. 11 is different from FIG. 11 in that processing for classifying all objects as flattening target / non-target (non-target) is performed (S301). Specifically, the estimated drawing time of each object is compared with the threshold value Sth, and the threshold value Sth or higher is classified as a flattening target, and less than the threshold value Sth is classified as non-flattening target (non-target).

  Note that the “expected drawing time” is, of course, considered that drawing is not actually performed at that time.

Thereafter, the processing is basically the same as the processing in FIG. 11, the window size A (2 ≦ A) in the processing number determination unit 24 is set to the initial value A1 (S302), and the N objects of the _Nth window A _N The overlap process (flattening process) and the N−1th rasterizer process (drawing process) of the object B _N−1 window are executed in parallel (S303a, S303b).

Then, to calculate the N-th window rasterizer processes the number _{B N} (S304), determines a flattening process number _{A N + 1} of the (N + 1) th window using the process number _{B N} (S305). In S305, A _{N + 1} is determined so as to satisfy the inequality (1). The above process is repeated until all objects are completed (S306).

  However, in the present embodiment, the rasterizer processing number B varies depending on conditions. That is, when the number of flattening processes is continuous, the number of flattening processes is M and B = M. On the other hand, when the number of flattening processes is skipped (when flattening target and non-target (non-target) are mixed), the sum of the number of non-target (non-target) until M counts is set as β, and B = M + β It is.

Specifically, it is as follows.
That is, there are objects 1 to 10 and the initial value A1 of the window size is set to A1 = 3. Compare the drawing time of each object with the threshold value Sth,
Object 1: Target Object 2: Target Object 3: Untargeted Object 4: Untargeted Object 5: Target Object 6: Target Object 7: Target Object 8: Target Object 9: Untargeted Object 10: Targeted .

  The flattening process targets are object 1, object 2, and object 5, and the flattening process is performed using these as windows. Then, the object 1, object 2, and object 5 subjected to the flattening process, and the five untargeted objects 3 and 4 between them are drawn. At the same time, the next processing number is determined using the inequality (1) using the processing number (processing time) of the objects 1 to 5. As a result, assuming that the number is determined to be four, flattening processing is performed using the object 6, object 7, object 8, and object 10 as windows. Further, the drawing process is executed with the objects 6 to 10 by adding the object 9 to these objects. In this embodiment, it should be noted that the flattening process is executed between objects having a relatively long drawing time.

Also in this embodiment, the applicant of the present application is
(I) 500 objects with 30% overlap in the page (ii) 1000 objects with 50% overlap in the page (iii) 500 objects with 70% overlap in the page In each case, the total processing time of the conventional method (the method shown in FIG. 2) is compared with the total processing time of the method of this embodiment, and it is confirmed that the following results are obtained.
In the case of (i) Conventional method: 1.4 seconds Embodiment method: 0.16 seconds (1/9 compared with the conventional method)
In the case of (ii) Conventional method: 5.2 seconds Embodiment method: 0.23 seconds (1/22 compared to the conventional method)
In the case of (iii) Conventional method: 126 seconds Embodiment method: 0.86 seconds (1/176 compared with the conventional method)
From these results, the speed of about 10 to 100 times is obtained in this embodiment compared with the conventional method, and the effect of shortening the processing time or increasing the processing speed according to this embodiment is clear.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to these embodiment, A various deformation | transformation is possible.

  For example, in the second embodiment, the number of objects in the window N + 1 is variably adjusted according to the drawing time of the window N. However, as a result of variably adjusting, the number of objects may be one. Therefore, in the first embodiment, the number of objects is an integer of 2 or more, but in the second embodiment, the number of objects is an integer of 1 or more and is not limited to 2 or more.

  In the third embodiment, the number of objects is variably adjusted in the same manner as in the second embodiment. However, the object number is more simply fixed (fixed to an integer of 1 or more), and the object is flattened. The flattening process may be performed by classifying the target / non-target (ratio target) and objects having a threshold value Sth or higher. That is, the third embodiment is not necessarily based on the configuration of the first embodiment or the second embodiment. Even in this case, the object that is not targeted for flattening is excluded from the flattening process, so that the processing time is shortened compared to the case where the entire flattening process is performed. Of course, in order to reliably hide the flattening process, it is desirable to determine the number of objects so as to satisfy the inequality (1).

  Furthermore, the image processing apparatus in each embodiment can be realized by causing a general-purpose computer to execute a program describing the processing of each functional module, for example. The computer includes, as hardware, a microprocessor such as a CPU, a memory such as a ROM and a RAM, various I / O (input / output) interfaces, a network interface that performs control for network connection such as a local area network, and the like.

20 Input / output control unit, 22 Object analysis unit, 23 Overlap omission target determination unit, 24 Processing number determination unit, 26 Overlap determination unit, 28 Overlap processing unit, 29 Next analysis number determination unit, 30 Processing window control unit, 32 Rasterizer drawing Part, 34 output control part, 36 reconstruction control part.

Claims (3)

Flattening processing means for performing flattening processing on image data by grouping image data number A (A is an integer of 2 or more);
Rasterizer processing means for generating a raster image from the flattened image data;
And processing the image data constituting the group N in the rasterizer processing means and the processing of the image data constituting the group N + 1 in the flattening processing means in parallel, where the group number is N (N is an integer of 1 or more) And executing the image processing apparatus. Flattening processing means for performing flattening processing on image data by grouping image data number A (A is an integer of 1 or more);
Rasterizer processing means for generating a raster image from the flattened image data;
And processing the image data constituting the group N in the rasterizer processing means and the processing of the image data constituting the group N + 1 in the flattening processing means in parallel, where the group number is N (N is an integer of 1 or more) And processing time of the image data constituting the group N + 1 in the flattening processing means according to the expected processing time prior to the processing in the rasterizer processing means of the image data constituting the group N. An image processing apparatus that variably sets the number A of image data constituting the group N + 1 so that the processing time of the image data constituting the group N in the rasterizer processing means is less than the processing time. Flattening that performs flattening processing on image data that is subject to flattening when the expected rendering time of the image data is relatively long, and flattening is not targeted when it is relatively short Processing means;
Rasterizer processing means for generating a raster image from the image data that has been flattened and the image data that has not been flattened;
An image processing apparatus that executes image data processing in the rasterizer processing means and image data processing in the flattening processing means in parallel.

Images