JP3414955B2 - Image synthesis processing device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image synthesizing processing device for synthesizing a plurality of images read by an image reading device and outputting them to an image recording device.

[0002]

2. Description of the Related Art An image forming apparatus receives image data of an original document read by an image reading apparatus and then subjected to predetermined image processing by an image processing apparatus, and based on the image data, an electrophotographic method or the like is used. Image formation is performed. Conventionally, as an image reading apparatus, Japanese Patent Laid-Open No. 4-192659
As presented in the above issue, a method of reading an image of a document using a two-dimensional area CCD has been proposed.

In this method, the original and the two-dimensional area CCD are not moved relative to each other during the reading of the image, and the original is collectively read on the reading surface of the two-dimensional area CCD whose relative position is fixed. Since the image is read, it may not be possible to read the entire image surface of the original at once in order to obtain high resolution for the image of the large original. Therefore, it is conceivable that the image of the original is divided into a plurality of parts and read, and then the plurality of images are combined and output.

Therefore, in the configuration disclosed in Japanese Patent Laid-Open No. 5-122606, a plurality of images to be combined are read so as to partially overlap each other, and a specific reference area is set within the overlapping range of one image. An area in which the difference between the reference area and the luminance value or color is 0 or the minimum is searched in the overlapping range of the other image, and both areas are overlapped to synthesize the image.

[0005]

However, in the configuration disclosed in Japanese Patent Laid-Open No. 5-122606, when there is no significant change in the luminance gradation in the reference area set within the overlapping range in one image, The corresponding region cannot be found in the other image, and the divided images cannot be accurately combined. In particular, if the reference area set in one image is small, it is highly likely that the wrong area in the other image will be determined as the area to be superimposed. Need a long time to. Further, since the sum of the differences between the pixels is calculated in the area comparison, when the minimum total value is checked while moving the area, the change in the total value due to the movement at the position where the total value is the minimum is If the error is slow and accumulated in each pixel due to the characteristics of the A / D conversion circuit of the light receiving unit, accurate comparison cannot be performed and a plurality of images cannot be accurately combined.

Such a problem is not limited to the case where the two-dimensional area CCD is used as the light receiving element, but the one-dimensional line C
The same occurs when combining a plurality of images read using a CD.

An object of the present invention is to provide an image synthesizing processing device capable of accurately synthesizing images divided into a plurality of parts regardless of the image content of a document.

[0008]

The invention according to claim 1
Are multiple images with overlapping areas that partially overlap each other
In the image synthesis processing device that synthesizes and outputs
Duplication of one image to be madeThe comparison target area is sequentially divided within the area.
The pixels of the comparison source area and the comparison destination area.
Correlation value map based on the sum of all squared degrees
And the minimum Euclidean distance from the correlation value map
Value S and the next smaller value, and the maximum Euclidean distance
The secondary differential value T at the small value position is extracted as a feature value,
Determines a search area that can be distinguished from other areas based on the features
Search area extraction processing, Duplication of the other image to be composited
Within the range, the composite area in which the feature quantity matches the search area is detected.
A synthetic area search process for searching, One image and the other
To the state where the position of the search area and the position of the composite area overlap.
It is characterized by executing an image synthesizing process for synthesizing
It

According to the first aspect of the invention, when two images are combined, first, the square of the brightness difference for each pixel.
Create a correlation value map based on the sum of
A search region that is distinguished from other regions based on the feature amount extracted from the image is extracted, and then a composite region whose feature amount matches the search region is searched for in the overlapping range of the other image. The search area is distinct from other areas in the overlapping area, and since the image content of the overlapping area in the other image matches the image content of the overlapping area in the one image, the search area in the overlapping area of the other image is Also, the search area is distinguished from other areas. Therefore, each pixel
Synthesis region feature amount extracted from the correlation value map based on the values match the search area is searched easily and accurately with distinguished from other regions in the overlap region of the other image, the synthesis of the two images exactly Done.

[0010]

[0011]

According to a second aspect of the present invention, there are provided first and second storage means for storing one image and the other image to be combined, respectively, and the first in the search area extracting process.
Read out the image from the storage means, execute the processing, read out the image from the first storage means and the second storage means in the composite area search processing, execute the processing, and based on the processing results, the plurality of storage means And an arithmetic means for synthesizing and outputting the images stored in.

According to the second aspect of the present invention, one image and the other image to be combined are stored in separate storage means, respectively, and the search area extraction is performed for the one image read from the first storage means. The calculation related to the process and the calculation related to the combined area search process for the other image read from the second storage unit are executed by the same calculation circuit. Therefore, the configuration of the device is simplified.

According to a third aspect of the present invention, a storage means for storing all of a plurality of images to be combined, and an arithmetic operation for storing one image and the other image read from the storage means for storage. After storing the storage means and all of the plurality of images to be combined in the storage storage means, the one and the other images are input from the storage storage means to the calculation storage means to perform the search area extraction processing, and Arithmetic means for executing the composite area search processing and combining and outputting images based on the processing result.

According to the third aspect of the invention, all of the plurality of images to be combined are temporarily stored in the storage means for storage, and the necessary images are read out from the storage means for storage to the storage means for calculation. Region extraction processing and synthetic area search processing are executed. Therefore, when the interrupted synthesizing work is restarted or the storage order of the images in the storage means for storage is changed very easily.

According to a fourth aspect of the present invention, there is provided an image storage means for storing either one image related to the search area extraction processing or the other image related to the composite area search processing, and the search area extraction means. Area storage means for storing the extracted search image, the search area extraction processing for one image stored in the image storage means, and the other image and the area storage means stored in the image storage means An arithmetic means for executing the composite area search process based on a search area, outputting the one image after the search area extraction process is completed, and outputting the other image after the composite area search process is completed. And are provided.

According to the fourth aspect of the present invention, the search area extraction processing and the composite area search processing are executed every time one image is stored in the image storage means, and the extracted search area is stored in the area storage means. The image stored in the image storage means is output. Therefore, by repeating this process, only the image storage means for storing one image and the area storage means for storing the search area are required, and the storage capacity to be prepared is reduced.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram showing a schematic configuration of an image composition processing apparatus which is an example of an embodiment of the present invention. The image composition processing apparatus 1 includes a document table 110, a lens 120, a light receiving section 130 including a two-dimensional area CCD sensor, a halogen lamp 80, a moving motor 90, an amplifier 140, an A / D converter 150, and a filter circuit 16.
0, field memories 10 and 11, address generator 20
-22, selectors 30-32, Euclidean calculator 4
0, the correlation value map memory 12, the correlation value map feature amount generator 50, and the controller 60.

The halogen lamp 80 irradiates the original 100 placed on the original table 110 with light. Lens 12
0 indicates the light receiving unit 13 that reflects the reflected light on the image surface of the original 100.
Focus to 0. The moving motor 90 moves the halogen lamp 80 and the light receiving unit 130 in parallel to the image surface of the document 100 placed on the document table 110. Light receiving unit 130
Is moved to a plurality of stop positions on the lower surface of the document table 110 by driving the moving motor 90, and the image of the document is read at each of the plurality of stop positions. The amplifier 140 amplifies the output of the light receiving unit 130. The A / D converter 150 converts the analog signal output from the amplifier 140 into digital data. The filter circuit 160 subjects the output data of the A / D converter 150 to processing such as median filtering and edge enhancement.

The field memories 10 and 11 alternately store the output data of the filter circuit 160. That is, the images read by the light receiving unit 130 at each of the plurality of positions are alternately stored in the field memories 10 and 11. Therefore, the image stored in the field memory 10 and the image stored in the field memory 11 are combined. The selectors 30 and 31 distribute the outputs from the field memories 10 and 11 to the Euclidean arithmetic unit 40 according to instructions from the controller 60, and output them. Euclidean calculator 40 is selector 3
Euclidean operation is performed on the outputs from 0 and 31,
The result is output. The correlation value map memory 12 stores the output from the Euclidean calculator 40. The address generators 20 to 22 output the images read from the field memories 10 and 11 to the outside according to the instruction from the controller 60.

Below, an explanation will be given of the processing in the case of synthesizing two images in the image synthesizing processing apparatus 1 as an example. 2 is a main flowchart of the image composition processing apparatus, FIG. 3 is a flowchart relating to document reading processing, FIG. 4 is a flowchart related to search area extraction processing, FIG. 5 is a flowchart related to composition area search processing, and FIG. FIG. 6 is a schematic diagram of a correlation value map generated at the time of synthesis processing.

When synthesizing two images, the image synthesizing processor 1 reads the image of the first range of the original document with the halogen lamp 80 and the light receiving section 130 positioned at the first stop position. Processing is performed (s10), and arithmetic processing is performed to extract a search area serving as a reference when the images are combined in the first image (s11). Next, the moving motor 9
When 0, the halogen lamp 80 and the light receiving unit 130 are moved to the second stop position (s12), the reading process is performed on the image of the second range of the document (s13), and the second image is matched with the search area. Arithmetic processing for retrieving a power synthesis area is performed (s14). The first image and the second image are output to the outside based on the position of the searched composite area (s1
5).

Next, s10, s11, s13 and s14
The processing in each of the above will be described in more detail.

In the reading process of the first image in s10,
The controller 60 drives the moving motor 90 to move the halogen lamp 80 and the light receiving unit 130 to the first stop position, selects the first field memory 10 (s20), and drives the halogen lamp 80. Then the first of the manuscript
The image of is read (s21). The received light signal of the light receiving unit 130 is amplified by the amplifier 140 (s22), converted into digital data by the A / D converter 150 (s23), and filtered by the filter circuit 160 (s24), and then the selected first signal is received. The image data is stored in the field memory 10 (s25). The image data writing process to the first field memory 10 is performed line by line up to the last line while updating the write address in the field memory 10 (s26, s27). This s
Through the processing of 20 to s27, the controller 60 sets the image data read from the first image in the two image ranges of the image of the original to the first field memory 10
Save in.

In the search area extraction processing in s11,
The controller 60 first sets the selectors 30 and 31 in the same field memory comparison state (s30). This same field memory comparison state is a state in which image data in two areas in one field memory are compared. In the same field memory comparison state, each of the selectors 30 and 31 is connected to the first field memory 1
The image data read from 0 is used by the Euclidean calculator 40
To enter. Next, the controller 60 sets the address of the comparison source area and the address of the comparison target area in the address generator 20 (s31, s32).

As shown in FIG. 6, the size of the overlapping range 203 that should overlap with the second image range 204 in the first image range 200 is predetermined, and within this overlapping range 203, the comparison source is compared. Area 201 and comparison target area 2
02 is set. Further, the sizes of the comparison source region and the comparison target region are also predetermined, and the comparison source region 201 and the comparison target region 202 have coordinate points P0 (x0, y0) and coordinate points P1 (x1, respectively) in the field memory 10. y
Specified by 1). Therefore, the controller 60
The address corresponding to each of the coordinate points P0 and P1 is the address of the comparison source area 201 and the comparison destination area 20.
The address of 2 is set in the address generator 20.

As a result of the processing of s30 to s33, the image data included in each of the comparison source area 201 and the comparison destination area 202 in the first image range 200 stored as image data in the field memory 10 is It is input to the Euclidean arithmetic unit 40 via the selectors 30 and 31. The Euclidean arithmetic unit 40 compares the comparison target area 2 with respect to the image data included in the input comparison source area 201.
The operation of the following first equation for obtaining the Euclidean distance ED of the image data included in 02 is executed (s33).

[0028]

[Equation 1]

Here, L is the luminance or chromaticity which is the image data of each pixel, and m and n are x of the regions 201 and 202.
It is the number of pixels in the direction and the y direction. Therefore, the Euclidean calculator 40 squares the luminance difference or chromaticity difference between the image data included in the comparison target area 202 and the image data at the corresponding position in the comparison source area 201, and compares the obtained value with the comparison target area. The total sum of 202 is calculated as the Euclidean distance ED. The Euclidean calculator 40 uses the calculated Euclidean distance ED as the correlation value map memory 1
2 (s34).

The controller 60 controls the comparison target area 202 in the range excluding the comparison source area 201 in the overlapping area 203.
Are sequentially allocated, the Euclidean distance ED for each comparison target area 202 with respect to the comparison source area 201 is calculated, and the Euclidean distance ED is stored in the correlation value map memory 12 (s35 → s32).
As a result, the correlation value map 210 for the comparison source area 201 is generated in the correlation value map memory 12.

Next, the correlation value map feature quantity generator 50
Is the difference S between the minimum value and the next smallest value of the Euclidean distance ED forming the correlation value map 210, as the feature amount of the correlation value map 210 for the comparison source area 201, and
2 by x, y at the minimum value position of Euclidean distance ED
The secondary differential value T is extracted, and this value is output to the controller 60 (s36). The controller 60 confirms whether the comparison source region 201 can be distinguished from other regions based on the extracted feature amounts S and T (s37), and the feature amounts S and T are determined.
Is larger than the determinable level and the comparison source area 201 can be distinguished from other areas, this area is determined as a search area and the next processing is performed.

When the feature amounts S and T are smaller than the determinable level and the comparison source region 201 cannot be distinguished from other regions, the controller 60 determines that the comparison source region 2 has been set.
It is determined whether 01 is the final area in the overlapping area 203 (s38), and if the set comparison source area 201 is not the final area, another area in the overlapping area 203 is set as the comparison source area 201. Is set to create a correlation value map with each comparison target area and extract a feature amount (s38 → s3
1). When the set comparison source area 201 is the final area, the controller 60 determines that the search area does not exist and executes the error processing.

In the reading process of the second image in s13,
In s12, the controller 60 sets the second field memory 1 while the halogen lamp 80 and the light receiving unit 130 are moved to the second stop position by the moving motor 90.
1 is selected and the same processing as the reading processing of s10 shown in FIG. 3 is executed. By this processing, the controller 60
The image data read from the twelfth image of the two image ranges in the image of the original is stored in the second field memory 11.

In the search processing of the composite area in s14,
The controller 60 first sets the selectors 30 and 31 in the different field memory comparison state (s50). This difference field memory comparison state is a state in which image data in areas in two field memories are compared. In this difference field memory comparison state, the selector 30,
Each of 31 inputs the image data read from each of the first field memory 10 and the second field memory 11 to the Euclidean arithmetic unit 40. The controller 60 then controls the address generators 20 and 21.
The address of the comparison source area and the address of the comparison destination area are set in (s51, s52).

The address of the comparison source area set in s51 is the address of the search area calculated in the processing of s11. In addition, the address of the comparison target area set in s52 represents an area of a predetermined size within the overlapping range 206 that overlaps the first image in the second image 204 shown in FIG. 6 when the images are combined. Coordinate point P2 (x2, y2)
Is the address of.

The selector 30 reads the image data of the search area from the first field memory 10 and inputs it to the Euclidean calculator 40. On the other hand, the selector 31 is the second
The image data of the comparison target area is read from the field memory 1 and input to the Euclidean calculator 40. The Euclidean computing unit 40 obtains the Euclidean distance ED of the image data included in the comparison target area 205 with respect to the image data included in the comparison source area 201, which is the search area, by the same processing as s33 shown in FIG. 4 (s53). The calculated Euclidean distance ED is stored in the correlation value map memory 12 (s54).

The controller 60 sequentially allocates the comparison target areas 205 in the overlapping area 206, and compares the comparison target areas 205 with each other.
Euclidean distance E with respect to the comparison source region 201
D is calculated and stored in the correlation value map memory 12 (s
55 → s52). As a result, the correlation value map 211 for the comparison source area 201, which is the search area, is generated in the correlation value map memory 12.

The correlation value map feature amount generator 50 has a correlation value map 211 generated in the correlation value map memory 12.
At, the data having the minimum value of the Euclidean distance ED is searched (s56), and the address of the comparison target area having the minimum value of the Euclidean distance ED is output to the controller 60. Euclidean distance ED in correlation value map 211
It is considered that the comparison target area having the smallest value is a combined area that matches the search area (comparison source area) 201 of the first image 200. Therefore, the controller 60 is
The correction data Mx, My of the address of the image data stored in the field memory 11 is calculated by the formula.

[0039]

[Equation 2]

Here, (x0, y0) is the coordinates of the point representing the search area in the first image, and (x2, y
2) is the coordinates of a point representing the composite area in the second image. Correction data M obtained by the calculation according to the second equation
x and My are subtracted from the coordinate data corresponding to the address of the first field memory 10 or added to the coordinate data corresponding to the address of the second field memory 11 to obtain the second field memory 11 The address of the image data stored in the first field memory 10 is continuous with the address of the image data stored in the first field memory 10.

In the external output process in s15, the controller 60 outputs the image data stored in the first field memory 10 in the order of addresses, and then outputs the address of the last image data output from the first field memory 10. From the next address, the correction data Mx,
My is subtracted, and the corrected address is used as the address generator 21.
Set to. As a result, the selector 32 sequentially reads and outputs the image data in the second field memory 11 from the corrected address. By this processing, the image combining processing device 1 outputs the image data of the first image and the image data of the second image read from the image surface of the original 100 in a completely combined state.

As described above, the image composition processing apparatus 1
When synthesizing two images, first, within the overlapping range of the two images in one image, based on the correlation value (feature amount) that is the sum of the brightness differences for each pixel, Extract a region that can be distinguished from other regions as a search region,
Next, a synthesis area whose feature amount matches the search area is searched for within the overlapping range in the other image. For this reason,
The problem of the prior art, which may cause a mismatch due to the fixed area to be compared, can be corrected. Further, since the comparison is performed at the characteristic points, the comparison area can be made small, and when the comparison area is made large by the conventional technique to avoid the mismatch, the disadvantage that the gradation difference comparison time becomes long can be corrected.

FIG. 7 is a block diagram showing the arrangement of an image synthesis processing apparatus which is an example of the embodiment of the invention described in claim 3 . 7, parts that are the same as the parts shown in FIG. 1 are given the same reference numerals, and descriptions thereof will be omitted. The image composition processing device 2 includes a magnetic disk controller 170 and a magnetic disk storage device 180 in addition to the configuration shown in FIG. The magnetic disk controller 170 is connected to the field memories 10 and 11 and the controller 60, and writes / reads data to / from the magnetic disk of the magnetic disk storage device 180. The magnetic disk storage device 180 holds and erases data under the control of the magnetic disk controller 170.

FIG. 8 shows a main flow chart in the image composition processing apparatus according to this embodiment. It should be noted that the flowchart of the document reading process, the flowchart of the search region extracting process, the flowchart of the combining region searching process, and the schematic diagram of the correlation value map generated during the image combining process are the same as those in FIGS. 3 to 6. Is.

First, the reading process shown in FIG. 3 is executed for a plurality of images, and the image data stored in the field memories 10 and 11 is written to the magnetic disk device 180 via the magnetic disk controller 170 (s60 to s6).
3). When the reading processing of all the images to be combined is completed (s63), the image data stored in the magnetic disk storage device 180 is read into each of the field memories 10 and 11 via the magnetic disk controller 170 (s64). After that, the extraction processing of the search area shown in FIG. 4 and the search processing of the composite area shown in FIG. 5 are executed (s65, s66), and the image data is calculated based on the position in the overlapping range of the searched composite area. Is output (s67,
s68).

As described above, in the image synthesizing processing apparatus 2 according to this embodiment, since the synthesizing work including the search region extracting process is executed after reading a plurality of images to be synthesized, the synthesizing interrupted midway. It is possible to easily change the process when resuming the work and the process procedure when the image is read.

FIG. 9 is a block diagram showing the arrangement of an image synthesis processing apparatus which is an example of the embodiment of the invention described in claim 4 . 9, the same parts as those in FIG. 1 are designated by the same reference numerals and the description thereof will be omitted. The image composition processing device 3
A sample area storage memory 190 is provided in place of the field memory 11 having the configuration shown in FIG. The sample area storage memory 190 stores the image data of the search area extracted from the first image.

FIG. 10 is a main flow chart showing the processing procedure of the image synthesis processing apparatus according to this embodiment, FIG. 11 is a flow chart of the search area extraction processing of the image synthesis processing apparatus, and FIG. 12 is the image synthesis processing apparatus. It is a flowchart which concerns on the search process of a synthetic | combination area. The flowchart of the image reading process and the schematic diagram of the correlation value map generated during the image combining process are the same as those in FIGS. 3 and 6.

First, the reading process shown in FIG. 3 is performed on the first image (s70), and the search region extracting process is performed (s71 → s73). In this search area extraction processing, after performing the same processing as s30 to s38 shown in FIG.
The image data of the extracted search area is stored in the sample area storage memory 190 (s80). Next, the first image data stored in the field memory 10 is output to the external device (s74). After that, the reading process shown in FIG. 3 is performed on the second image to store the image data in the field memory 10 (s70), and the synthesis region search process is performed (s71 → s72). In the search processing of the composite area, the selector 30 is set in the sample area storage memory 190 (s90), and in this state, s51 to s shown in FIG.
The processing of 57 is executed. Further, the second image data stored in the field memory 10 is output to the external device based on the obtained correction data (s74).

When synthesizing three or more images, a search region extraction process is performed after the synthesis region search process for the second image transition image, and the images are output to the outside after the search process is completed. (S72 → s73 → s74).

With the above processing, the image composition processing apparatus 3 according to this embodiment is provided with the sample area storage memory 190 for storing only the search area, and after extracting the search area from the image, the image data is stored. The memory capacity can be reduced by outputting the data to the outside first and always storing only the image data read by one image reading process in the field memory 10.

In each of the above-described embodiments, the light receiving section 130 is configured by a two-dimensional area CCD sensor, and the image is read in a state where the light receiving section 130 does not move relative to the image surface of the document. However, even in the case where the light receiving unit 130 is configured by a one-dimensional line CCD sensor and the image is read while the light receiving unit 130 is moving relative to the image surface of the document, The invention can be practiced similarly. In this case, by replacing the original whose image surface is larger than the image reading range of one time by the reading unit 130 a plurality of times by changing the position facing the original plate in a partially overlapping state, the entire image surface of the original is replaced. The image data of can be accurately output.

In each of the above-described embodiments, the halogen lamp 80 for reading an image, the lens 120, and the light receiving unit 130 are provided, but a plurality of images other than these are input from an external device. It can also be configured as follows. Further, the image composition processing apparatus of the present invention can be arranged in a part of an image forming apparatus such as a copying machine.

Furthermore, in each of the above-described embodiments, the case where two images are combined has been described, but the present invention can be similarly implemented when combining three or more images. For example, in the image synthesizing processing device 1 shown in FIG. 1, three or more images are obtained by repeatedly performing the processing of s11 to s15 shown in FIG. 2 by the number of times 1 is subtracted from the number of images to be synthesized. Can be synthesized accurately. Further, as the number of images to be combined increases, the possibility that the combining operation will be interrupted in the middle increases, but in the image combining processing device 2 shown in FIG. 7, the processing of s60 to s63 shown in FIG. By repeating the above process the same number of times as the number of images to be combined, it is possible to easily perform the process of restarting the interrupted combining operation and the change of the process procedure at the time of image reading. Furthermore, FIG.
In the image composition processing device 3 shown in FIG.
By repeating the processing from 0 to s74 the same number of times as the number of images to be combined, the memory capacity to be prepared is remarkably increased as compared with the case of storing all of the image data of a plurality of images to be combined. Can be reduced.

[0055]

According to the first aspect of the present invention, when two images are combined, first, the square of the luminance difference for each pixel is used.
Create a correlation value map based on the sum of
A search region that is distinguished from other regions based on the feature amount extracted from the image is extracted, and then a composite region whose feature amount matches the search region is searched for in the overlapping range of the other image. The search area is distinct from other areas in the overlapping area, and since the image content of the overlapping area in the other image matches the image content of the overlapping area in the one image, the search area in the overlapping area of the other image is Also, the search area is distinguished from other areas. Therefore, each pixel
It is possible to easily and accurately search for a composite area in which the feature amount extracted from the correlation value map based on the value of the search area coincides with the search area from other areas in the overlapping range of the other image, and to accurately combine the two images. It can be carried out.

[0056]

According to the second aspect of the invention, one image to be combined and the other image are stored in separate storage means, respectively, and the search area for the one image read from the first storage means. The same arithmetic circuit can execute the arithmetic operation related to the extraction processing and the arithmetic operation related to the combined area search processing for the other image read from the second storage unit. Therefore, the configuration of the device can be simplified.

According to the invention described in claim 3 , all of the plurality of images to be combined are temporarily stored in the storage means for storage,
A necessary image is read out from the storage means for storage into the storage means for calculation, and a search area extraction process and a composite area search process are executed. Therefore, when restarting the interrupted synthesizing operation or changing the storage order of the images in the storage means for storage, it can be performed very easily.

According to the fourth aspect of the present invention, the search area extraction processing and the composite area search processing are executed every time one image is stored in the image storage means, and the extracted search area is stored in the area storage means. And the image stored in the image storage means is output. Therefore, by repeating this process, it suffices to provide the image storage means for storing one image and the area storage means for storing the search area,
The storage capacity to be prepared can be reduced.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of an image composition processing apparatus that is an example of an embodiment of the present invention.

FIG. 2 is a main flowchart of the image composition processing apparatus.

FIG. 3 is a flowchart relating to a document reading process in the image composition processing apparatus.

FIG. 4 is a flowchart relating to a search region extraction process in the image composition processing device.

FIG. 5 is a flowchart relating to a search process for a combined area in the image combining processing apparatus.

FIG. 6 is a schematic diagram of a correlation value map generated at the time of image synthesis processing.

FIG. 7 is a block diagram showing a configuration of an image synthesis processing apparatus which is an example of an embodiment of the invention described in claim 3 ;

FIG. 8 shows a main flowchart in the image composition processing apparatus.

FIG. 9 is a block diagram showing a configuration of an image synthesis processing apparatus which is an example of an embodiment of the invention described in claim 4 ;

FIG. 10 is a main flowchart showing a processing procedure of the image composition processing apparatus.

FIG. 11 is a flowchart relating to a search region extraction process of the image composition processing device.

FIG. 12 is a flowchart relating to a search process for a combined area of the image combining processing apparatus.

[Explanation of symbols] 1, 2, 3-image synthesis processing device 10, 11-field memory 20, 21-Address generator 40-Euclidean arithmetic unit

Claims (4)

(57) [Claims] 1. A compound having an overlapping range partially overlapping each other.
To an image synthesis processing device that synthesizes and outputs a number of images.
And Duplication of one image to be compositedCompare target areas sequentially within the area
Allocation, for each pixel of the comparison source area and the comparison destination area
The brightness difference is squared, and the correlation value map is calculated based on the total sum of the whole.
Of the Euclidean distance from the correlation value map.
The difference S between the small value and the next smallest value, and the Euclidean distance
The second derivative value T at the minimum value position is extracted as a feature quantity,
The search area that can be distinguished from other areas is determined based on the feature amount of
FixedSearch area extraction processing, The feature is searched for within the overlapping range of the other image to be combined.
Synthetic region search processing to search for a synthetic region that matches the region
When, The position of the search area and the combined area of one image and the other image
Image combining processing to combine the position of and
Image composition characterized by executionprocessingapparatus. 2. One of the images to be combined and the other of the images to be combined.
First and second storage means for storing each, and the search
An image is read from the first storage means in the area extraction processing.
Then, the process is executed, and
The image is read from the storage means and the second storage means and processed.
And executes the processing in multiple storage means based on the processing result.
The image composition processing apparatus according to claim 1 , further comprising: an arithmetic unit configured to combine and output the stored images . 3. A memory for storing all of a plurality of images to be combined.
One of the existing storage means and the one read from the storage means for storage.
An arithmetic memory for storing each of the image and the other image
And a plurality of images to be combined in the storage means for storage
After saving, save the one and the other images
Input to the storage means for calculation from the stage
Management, and the combining region search processing is executed, the image synthesizing apparatus according to claim 1 in which the calculating means, provided for outputting the synthesized images on the basis of the processing result. 4. One of the images related to the search area extraction processing.
Or one of the other images related to the composite area search processing
The image storage means for storing
Area storage means for storing the issued search image, and image storage
Extraction of the search area for one image stored in the means
Processing and the other image stored in the image storage means
The composite area based on the search area stored in the area storage means
Area search processing and search area extraction processing
Output the one image after completion of the
The image synthesizing processing device according to claim 1, further comprising a calculation unit that outputs the other image after the search process is completed .