JPH0556251A - Image reader - Google Patents

Abstract

PURPOSE:To improve the processing speed and the quality of an image by omitting the distortion correcting processing of the divided image discriminated to be that the whole surface is the uniform concentration, processing the junction synthesization of the divided image from a central part to a peripheral part and decreasing the burden of an image processing circuit. CONSTITUTION:When the desired resolution is instructed from an input instructing part 11 and the reading of an original image 14 is instructed, an image pick-up scanning part 13 picks up the image. At the time of picking up the whole image, the edge of the image 14 is detected, the distance and angle information of the scanning part 13 and the image 14 are obtained by focusing and the size of the image 14 is operated. An image processing circuit 122 calculates the resolution and zoom quantity, and by the result, sets an image forming optical system 133. The circuit 121 calculates the number of image divisions and performs the image division scanning. When a discriminating circuit 123 discriminates that the whole surface is the uniform image, the distortion correcting processing of the divided image is omitted. When it is discriminated that it is not uniform, the distortion correction is performed. Next, the junction synthesization of the divided image 14 is processed from a central part to a peripheral part and an adjoining area is picked up so that several picture elements can be overlapped.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading apparatus capable of reading an original image placed on an unspecified place, and is used for an image reading apparatus such as a facsimile, a digital copying machine, a word processor, or a computer. It is possible.

[0002]

2. Description of the Related Art An image reading apparatus for reading an image of a document placed on a specific place in a conventional example will be described with reference to FIGS. FIG. 12 shows an image reading apparatus when the optical system is moved and the solid-state image sensor is fixed. FIG. 13 shows an image reading apparatus when the optical system and the solid-state image sensor are moved. In FIG. 12, the reading device main body 91 includes a platen 93 on which a document 92 is placed and a document 92.
An illumination light source 94 for irradiating a linear light source on the light source, mirrors 95-1 to 95-3 forming an optical system for transmitting the light reflected from the original 92 to a solid-state image sensor 97 described later, and an image of the original 92. The solid-state image sensor 97 includes a lens assembly 96 for forming an image on the solid-state image sensor 97, and a solid-state image sensor 97 for capturing an image of the original 92 and converting the image into an electric signal.
Then, the illumination light source 94 and the mirrors 95-1 to 95-3
Can be scanned along the original 92 by a driving device (not shown). In such an image reading apparatus, when the document 92 placed on the platen 93 is irradiated with the linear illumination light source 94, the reflected light reflected from the image of the document 92 is reflected by the mirrors 95-1 to 95-3 and An image is formed on the solid-state image sensor 97 via the lens assembly 96. Then, the image of the original 92 is scanned by the illumination light source 94 and the mirrors 95-1 to 95-3 from the left to the right in the figure so that the solid-state image sensor 97 reads the entire surface of the original 92.

In FIG. 13, what is different from FIG. 12 is the reading section of the original 92. That is, the reading unit is composed of the linear illumination light source 94, the rod lens array 98, and the solid-state imaging device 97, and moves along the lower portion of the platen 93. In such an image reading apparatus, when the document 92 placed on the platen 93 is irradiated with the linear illumination light source 94, the reflected light reflected from the image of the document 92 passes through the rod lens array 98 to perform solid-state imaging. An image is formed on the element 97. Then, the image of the original 92 is read by the solid-state imaging device 97 over the entire surface of the original 92 by scanning the reading section from left to right in the figure. However, in the conventional example shown in FIGS. 12 and 13, a platen for reading an original is always required, and the original must be smaller than the size of the platen. Also, when placing the original on the platen, the original is placed on the back side.
Positioning was difficult when the image drawn on the original was placed at a desired position and imaged. Further, since the scanning unit and the optical system have the driving unit for scanning the entire platen, the size of the image reading apparatus becomes very large. In the stationary image reading apparatus as in the above conventional example, when the image to be read is large and heavy, it cannot be placed on the platen. There is a handy scanner as a solution to the size of the stationary image reading device, but there is a problem that scanning accuracy is poor because the document is traced by hand and it cannot be used for commercial use. did.

In order to solve the above problems, the present applicant has
An image reading apparatus that obtains one document image data by swinging a unit containing a solid-state image sensor vertically and horizontally to divide and capture the document image, and then perform the splicing and combining process of the divided images. Proposed. For example, the image reading device is dated September 18, 1990 (Japanese Patent Application No.
No. 246291) has been filed.

A document 92 is placed, for example, a platen 9
An image reading apparatus that does not require a specific mounting table such as No. 3 will be described with reference to FIGS. 14 and 15. Figure 1
4 is a schematic diagram of an image reading apparatus in a conventional example. Figure 1
FIG. 5 is a diagram illustrating a scanning unit of the image reading apparatus in FIG. In FIG. 14, the document 251 is placed on, for example, an office desk, not on a specific placing table. The original 251 is divided into a desired number of divided images 252 by an image reading device described later, and each divided image 252 is read. Such an image reading apparatus is capable of scanning in the X-axis and Y-axis directions, and has a scanning unit 253 having a zoom mechanism, a supporting unit 254 that supports the scanning unit 253, and a stand base 255 that supports the supporting unit 254. And an image processing device (not shown). Hereinafter, the scanning unit 253 will be described in detail.

In FIG. 15, the scanning mechanism section 201 includes a support section 254 to which a horizontally movable housing 203 is attached, and a horizontal scanning mechanism section 209 in which the horizontally movable housing 203 is rotatably installed around a horizontal scanning axis 207. , Horizontally movable housing 2
No. 03, which is installed in the vertical movable mechanism 211 and is rotatable around the vertical scanning axis 221 by the vertical scanning mechanism 223, and a magnification control mechanism 217 provided in the vertically movable casing 211.
Also, it is composed of an image forming optical means 213 for forming an image of a document by an optical means such as a focus control mechanism 219, a solid-state image pickup device 215 for reading and a magnification control mechanism 217 which are also provided on the vertically movable housing 211.

The scanning mechanism section 201 having the above-mentioned configuration is
The scanning of the original is started in accordance with a desired resolution of the image and a reading start instruction from an input instruction unit (not shown). An image of the entire original is read by this scanning. The edge of the document image is detected from the entire image of the document image. Further, the distance and the angle between the scanning unit and the document are calculated by focusing on the document. At the same time, the inclination of the document is calculated. Further, the zoom amount is calculated according to the designation of the resolution. Then, the number of divisions of the document image is calculated from the above calculation results.
The divided original image divided by this calculation is scanned in the X-axis and Y-axis directions by driving the scanning mechanism unit 201, and as a result, is read by the solid-state image sensor 215 through the imaging optical unit 213. The data read by the solid-state image sensor 215 is corrected by the correction circuit, because the original image is tilted, and the distortion caused by the difference in focal length occurs. The divided images corrected as described above are joined and combined to be assembled into one original image.

Techniques relating to the above-mentioned divided image processing, image distortion correction, oblique image correction, image focus adjustment, etc. are known from, for example, the following publication. That is, JP-A-59-45576 and JP-A-60-9250.
No. 7 JP-A-60-163182 JP-A-60-1631
No. 83 Japanese Patent Laid-Open No. 63-42442 Japanese Patent Laid-Open No. 63-3086
No. 78, JP-A-63-314970 and the like.

[0009]

However, in the above-mentioned image processing, the divided images have the disadvantages that the weight of each combined image is large and the processing time is long. Further, the splicing and combining process of the divided images is complicated and makes the price of the image reading apparatus expensive.

When the divided images are combined and combined, the density and distortion of the combined and combined images must be corrected. However, when the combined images are combined in the scanning order, the correction can be made between two adjacent images. However, correct correction is not always performed between the images in the vertical direction. In particular, among the corrections of adjacent (vertical and horizontal directions), if the preceding correction is prioritized, there is a problem in that the degree of correction to be performed later must be compromised.

Diffuse illumination was used as the illumination for illuminating the original image in order to prevent uneven illuminance.
It was expensive. Further, in the image reading apparatus of the type in which the divided images are combined and combined, the density difference is large due to the variation of the illuminance, and when the plural combined divided images are combined and combined, the correction in the gradation direction is required. Further, since the imaging scanning unit scans the document image from diagonally above, the geometrical joining synthesis for removing the distortion caused by this scanning has a problem that the processing load is large.

When a large original is read by the image reading apparatus as shown in FIG. 14, the resolution of the portion away from the stand is lowered. In order to prevent this reduction in resolution, the document surface is divided, but a step as shown in FIG. 16 is generated at the boundary portion formed by the splicing and combining process for combining these divided images into one. Especially when a character image is taken,
The step generated at the boundary is conspicuous, and it can be seen that the image is clearly deteriorated. Therefore, there is a problem in that processing must be performed so that the step generated at the boundary of the divided images is inconspicuous.

The present invention is intended to solve the above problems, and a divided image which does not require distortion correction processing is
An object of the present invention is to provide an image reading apparatus that omits the distortion correction processing to reduce the processing load and also improves the processing speed.

According to the present invention, the splicing synthesis is preferentially started from the vicinity of the central portion where the images are concentrated, and the unsatisfactory point of the splicing synthesis processing in the peripheral portion of the image is compromised to improve the image quality of the central portion of the image. An object is to provide an improved image reading device.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an image reading apparatus provided with a splicing / combining means that can easily superimpose superposed portions when splicing and combining divided images and improves image quality.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an image reading apparatus which, when joining and combining divided images, performs processing in the order in which the density changes a lot and the joining and combining are easy.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an image reading apparatus in which, when splicing and combining divided images, inconsistency in splicing and combining at the boundary is not noticeable.

[0018]

(First Invention) In order to achieve the above-mentioned object, an image reading apparatus of the present invention includes an image forming optical unit capable of changing a magnification and an image forming optical unit. And an imaging unit arranged in a predetermined positional relationship with each other, the imaging optical unit and the imaging unit can scan the original image in the X-axis and Y-axis directions, and can scan the original image by dividing it into a plurality of images. Means for combining and combining the divided and captured divided images to obtain one original image data. The density of the divided images is calculated, and the entire surface is uniform. The density determination means (123 in FIG. 1) for determining whether or not the density is uniform, and the distortion correction processing of the divided image determined by the density determination means (123) to have uniform density on the entire surface are omitted. Distortion correction processing means (12 in FIG. 1) ) Made from a.

(Second Invention) In the image reading apparatus of the present invention, the splicing and combining of the divided images are carried out in an order from the central portion to the peripheral portion.

(Third Invention) In the image reading apparatus of the present invention, the adjacent areas when the divided images are joined and combined are preliminarily picked up by the image pickup means so as to overlap by several pixels.

(Fourth invention) In the image reading apparatus of the present invention, the divided images are combined and combined in the descending order of the dispersion value of the density change along the edges of adjacent small areas of the divided images. It

(Fifth Invention) In the stitching synthesis of divided images in the image reading apparatus of the present invention, when the stitching synthesis of the divided images is performed, a pixel in the vicinity of a target pixel in an adjacent region is extracted, and pixels in the vicinity are extracted. Is calculated, and the average density is used as the density of the pixel of interest to be the image output signal of the adjacent region.

[0023]

[Operation] (First invention) The original image is captured as a whole by instructing the resolution and the start of reading, and the number of divisions of the image is determined by, for example, the edge, tilt, or zoom amount of the original. To be done. Next, each of the divided images is scanned and read in the X-axis and Y-axis directions by the image forming optical unit and the image pickup unit. The divided original image is geometrically corrected for an inclination caused by scanning from an obliquely upper direction or an inclination of an image caused by an inclination caused by placing the original. And
If necessary, for example, the density correcting unit corrects the original images divided in the scanning order so as to have the same density characteristic when the divided images are joined and combined. However, most of the original images are document images, and in most cases, there is a margin around the document image or a part thereof.
In addition, even in images such as pictures and graphs other than documents, the images are often concentrated in the central part, and the periphery thereof is often the background with a uniform density. For example, a plurality of divided images in which the entire surface is a white background or a plurality of divided images in which the entire surface has a uniform density appear with high probability. Therefore, when calculating the densities of the divided images, the density determining means determines that the entire surface of the divided images has a uniform density, for example,
It is determined whether the background is all white or the same color. Then, the distortion correction processing means omits the distortion correction processing of the divided image which is determined by the density determination means to have uniform density over the entire surface. In the image reading apparatus as described above, compared with the conventional example in which the distortion correction processing is uniformly applied to all the divided images, the processing is reduced by the amount by which the processing can be omitted and the processing speed is increased. be able to.

(Second Invention) The second invention focuses on the fact that normal images are concentrated in the central portion rather than the peripheral portion. That is, when the divided images are joined and combined, the distortion between the adjacent divided images is corrected,
The correction distortions will overlap later. For this reason, the image reading apparatus of the present invention starts joining and combining from the vicinity of the central portion where the image is concentrated, and compromises that the correction distortion generated in the peripheral portion of the image becomes large. In such a splicing and combining of divided images, even if distortion remains in the periphery of the image, if the image is small in the peripheral portion or concentrated in the central portion,
Satisfactory one image is obtained.

(Third Invention) When the divided images are combined and combined, the difference in density is utilized so that there is no difference in density at the edges of the divided images. In the third invention, the overlapping margin of the divided images is determined to be, for example, 2 to 3 pixels. 2 to 3 more than the case of joining and combining by matching the images
It is possible to perform the joining synthesis with higher accuracy when the pixels are matched with each other.

(Fourth Invention) According to the fourth invention, when the divided images are combined and combined, if the density difference is small, a slight deviation may be allowed, and the part having a large dispersion value of the density change is prioritized. Then, they were joined and synthesized. That is, the density determination means examines the variance value of the density change along the edge of the joined portion of the divided images, and since the density change is large, the portion in which the deviation is a problem is preferentially joined and combined. Since the deviation is less noticeable in the portion where the density change is small, the deviation may be large in this portion.

(Fifth Invention) According to a fifth invention, when a plurality of divided images picked up by division are combined and combined, a pixel in the vicinity of the pixel of interest in the combined combination area is extracted. For example,
It is assumed that nine pixels including the target pixel are extracted from the character image. In the case of a character image, the density of each pixel is white or black, so if there are, for example, five pixels on which characters are drawn among the nine extracted pixels, the density of the pixel of interest is The average value of the pixels in the vicinity is 5/9. That is, for example, if the pixel of interest has a density of "0" for white and a density of "1" for black, the pixel of interest is blurred to black with a density of 5/9. In this way, the densities of the target pixel in the junction combining area are sequentially calculated. Then, the density of each pixel in the junction composition area is output as the density of the average value based on the calculation result. As described above, since the densities in the splicing composite area are averaged and blurred, even if there is a density change in that portion, the step due to the density difference becomes an image in which the step is inconspicuous as an output image. ..

As described above, since the image reading apparatus of the present invention inputs the original image by the image forming optical system and the two-dimensional image input device such as the solid-state image pickup device, the original is placed at a specific place such as the platen. It is not necessary to place it, and it can be read regardless of the arrangement state or shape of the document. Further, in the image reading apparatus of the present invention, the processing that does not need to perform the joining synthesis or the distortion correction is omitted, the processing speed is increased, the place where the amount of deviation is conspicuous is given priority, and the place where the deviation is conspicuous is left behind. By processing, the quality can be improved as one image.

[0029]

[Embodiment] FIG. 1 is a block diagram for explaining an embodiment of the present invention. In FIG. 1, the image reading apparatus includes an input instruction unit 11, an image processing unit 12, and an imaging scanning unit 13.
Composed of and. The input instructing section 11 is composed of a resolution specifying section 111 for specifying a desired resolution and a reading start specifying section 112. Image processing unit 12
Is a control circuit 121 that controls the entire image processing unit 12,
The image processing circuit 122 is configured to divide the original image and read the divided image, or to combine and combine the read divided images. The image processing circuit 122 includes a density determination circuit 123 that determines the density of the divided images, and a density correction circuit 124 that corrects the density in order to make the gradation of each divided image as constant as possible when the divided images are combined and combined. If all the densities in the divided images are uniform, the distortion correction processing circuit 1 omitting the processing for correcting the distortion when capturing the original image
25, a joining combination order determination circuit 126 that determines the joining order when joining the divided images in a desired order, and a density average value calculation circuit 127 that calculates an average value of the densities in the joining combination areas of the divided images. To be done. Imaging scanning unit 13
Is a drive processing circuit 131 that freely scans the imaging scanning unit 13 in the X-axis and Y-axis directions, an image sensor 132 that reads a document image, for example, a MOS type, a bipolar type, or a TFT type solid-state image sensor, and a document. An imaging optical system 133 for reading the image 14 and focusing it on the image sensor 132, a drive mechanism 134 for scanning the imaging optical system 133 in the X-axis and Y-axis directions, and the drive. The control circuit 135 controls the mechanism 134.

A desired resolution is designated from the input designating section 11 and at the same time, an instruction to read a document image is given. The above two instructions are input to the control circuit 121 of the image processing unit 12. The control circuit 121 sends a control signal to the drive processing circuit 131 of the imaging scanning unit 13 and the control circuit 135 of the imaging scanning unit 13 according to an instruction from the input instructing unit 11. The imaging optical system 133 operates the drive mechanism 134 under the control of the control circuit 135 to adjust the original image 14 so that it is focused on the image sensor 132, or change the zoom amount. Further, the image data of the original image read by the image pickup device 132 is sent to the image processing unit 12 in accordance with the timing of the control signal of the drive processing circuit 131. Further, the image processing circuit 122 of the image processing unit 12
The original image data is obtained after performing a predetermined process by.

Next, the image processing in the image processing circuit 122 will be described in detail with reference to FIG. FIG. 2 shows a flow chart for explaining an embodiment of the present invention.
FIG. 3 is an explanatory diagram when joining and combining divided images. At the same time as instructing a desired resolution from the input instructing unit 11, it is instructed to start reading the original image 14 (steps 21, 2).
2). The imaging scanning unit 1 is instructed by the instruction of the input instruction unit 11.
3 captures the entire document image 14 (step 2
3). Imaging of the entire image is also called pre-scan, and the operator may manually read the display while changing the zoom amount while looking at the display. When the entire original image 14 is automatically imaged, characteristic points of the original image 14, such as edges, are detected (step 24). The edge is detected by, for example, detecting the difference in contrast between the color of the upper surface of the office desk on which the document is placed and the color of the document. Therefore, when the original is white paper, it is desirable that the color of the upper surface of the desk is not white. Further, adjustment is performed in order to focus when capturing the entire original image 14. By this focusing, the distance and angle information between the imaging scanning unit 13 and the document image 14 can be obtained. The image processing circuit 122 calculates the size of the document image 14 from the edge of the document image 14 and the distance and angle information (step 25). For example,
It can be seen whether the size of the original image 14 is A4 or B5. If the original image 14 has no inclination and the resolution does not matter in advance, steps 26 and 27 to be described later are skipped and the process proceeds to step 28.

The image processing circuit 122 calculates the zoom amount from the desired resolution designated by the input designating section 11 and the original image 14, and the imaging optical system 133 is calculated based on the result of the calculation.
Is set (step 27). Even with the same original image 14,
If the place where the original image 14 is placed changes, the zoom amount also changes.
If it is known in advance that the size and the original image 14 are not tilted before reading the original image 14, it is possible to skip steps 22 to 26 and start the calculation from step 27. Image processing circuit 122
Calculates the number of divisions of the document image 14 according to the size and zoom amount of the document image 14 (step 28). When the number of divided original images 14 is determined, the size of each divided image is determined. Therefore, each divided image is divided into, for example, A, B, ... As shown by arrows in FIG.
All original images are scanned in the order of O and P (step 29).
When the original image 14 may be read as one image regardless of the resolution, steps 27 to 29 can be skipped and the process can proceed to step 30. If the original image 14 is not tilted and the whole image that is not divided is acceptable, step 26 to step 2
Skip 9 and go to step 30.

The image processing circuit 122 corrects the gradation of the divided image (step 30). Then, the density determination circuit 123 reads the divided image and determines whether or not the entire surface is a uniform image (step 31). For example, when it is determined that the entire surface of the divided image is a white background, all the output signals of the density determination circuit 123 are “0”. When it is determined that all the output signals are "0", the next process, that is, the distortion correction process of step 32 is omitted,
The process proceeds to step 33 where the divided image combining process is performed. Also,
The same determination can be performed when the color of the divided image is gray or yellow other than white, or when the image is a picture. When it is determined in step 31 that the entire surface of the divided image is not uniform, that is, when the output signal of the density determination circuit 123 has "0" and "1", the distortion when the image is captured obliquely from above is geometrically determined. Then, the process for correcting is performed (step 32). Then, the image processing circuit 122
The divided images are joined and combined by (step 3).
3). For joining and combining the divided images, for example, the linearity of the line drawing is used, and the joining and combining are performed so that the portions match. The above divided images are continuously processed until there is no object to be joined and combined (step 34). When there are no divided images to be joined, the data of the original image is output under the control of the control circuit 121 of the image processing unit 12. Note that the original image 1
After dividing all four, the original image 14 is divided into one and then step 3
After performing the processes of 0 to 33, it is possible to return to step 28 (dotted line arrow in FIG. 2) and repeat this for all divided images.

When the original image is read by the above-described embodiment, as shown in FIG. 3, for example, white divided images D, E, J, L, M, and P are subjected to distortion correction processing and splicing synthesis processing. You don't have to. Therefore, it is possible to provide an image reading apparatus with a short processing time.

Next, another embodiment of the present invention will be described with reference to FIG. FIG. 4 is a flowchart in the case where the joining and combining order of divided images is started from the vicinity of the center. FIG. 5 is an explanatory diagram in which the splicing and combining order of divided images is started from the vicinity of the center. In FIG. 4, the processing from step 21 to step 29 is the same as that shown in FIG. The present embodiment focuses on the fact that images are generally concentrated near the central portion of the document image 14. Since the imaging scanning unit 13 of the present embodiment captures a document image from diagonally above, there is a scanning angle of the imaging scanning unit 13 with respect to the document image 14, and the angle is different for each divided image. When there is an error in the scanning angle between the adjacent divided images when the divided images are combined, even if the error is a slight error, it will be displaced by several pixels at the joint portion. Therefore, black and white matching correlation is obtained at the edge portion of the area centered on the joining point of adjacent divided images, and processing is performed for joining at the best matching location. For this reason, each of the divided images is provided with an overlapping portion 51 for several pixels in advance.

Next, step 40 of the flow chart in FIG. 4 will be described. The divided image scanned by the imaging scanning unit 13 from diagonally above has more distortion than the one scanned from directly above. The distortion of this divided image is corrected geometrically (step 40). Then, the corrected divided images are joined next to each other. As mentioned above, since there are many line drawings in the center of the image and it is easy to extract feature points, it is easy to join, and in the peripheral part of the image, there are many parts without images.
Joining tends to be difficult. In addition, when adjoining images are joined together, even if there is a slight deviation due to an error in correcting the distortion, if such divided images are successively joined, the amount of the deviation gradually increases. Therefore, the joining / combining process of the divided images is started from the vicinity of the center of the original image without performing the joining / combining from the edge of the original image as in the conventional example. As for the sequence of the joining and combining process of the divided images, for example, as shown by a solid arrow in FIG. 5, a process of starting from 1 and circling spirally to reach 20 is performed. Further, in FIG. 5, the joining and combining order is 2, 3, 1, which is the center of the image.
After setting 4, 6, and 5, the peripheral portion of the image may be circulated.
Further, when the important part of the image is biased or dispersed, the part may be preferentially combined. The order of such divided image joining synthesis is
In the joining / combining order determination circuit 126 in the image processing unit 12, the input instruction unit 1 has a program in advance.
The order can also be determined by the instruction of 1. As described above, in the present embodiment, the important part of the image is preferentially subjected to the splicing synthesis, and the distortion of the less important part of the image is allowed, so that the quality of the entire image is improved.

Next, another embodiment of the present invention will be described with reference to FIGS. FIG. 6 is a flowchart in the case where the joining and combining order of divided images is started in the descending order of the density dispersion value. FIG. 7 is a diagram illustrating before and after joining of divided images with high density. FIG. 8 is a diagram illustrating before and after joining of divided images with low density. In FIG. 6, the processing from step 21 to step 29 is the same as that shown in FIG. 2 and FIG. Generally, it is possible to carry out splicing and combining with a small error in a portion where the density change of an image is large, and it is difficult to splicing and combining images in a peripheral portion of an image with many white backgrounds. Since the imaging scanning unit 13 of the present embodiment captures a document image from diagonally above, there is a scanning angle of the imaging scanning unit 13 with respect to the document image 14, and the angle is different for each divided image. Therefore, when the divided images are combined, if the adjacent divided images have distortion, the distortion tends to increase each time the divided images are combined. Therefore, if the density change in the divided image composition unit is large,
Since the feature points are easily extracted, the joining can be easily performed, and when the density change in the divided image combining unit is small, it tends to be difficult to extract the feature points. Further, when the change in the concentration is small in the joining / synthesizing portion, even if there is some distortion, it is not very noticeable.

Next, step 60 of the flow chart in FIG. 6 will be described. Similar to step 40 shown in FIG. 4, the divided image scanned by the imaging scanning unit 13 from diagonally above has more distortion than that scanned from above. The distortion of this divided image is corrected geometrically (step 60). When synthesizing the divided images, the variance value of the density change of the splicing synthesizing part is checked (step 61). Then, the descending order of the density dispersion value of the divided images is determined by calculation. The order determined by this calculation becomes the joining synthesis order of the divided images (step 62). Then, the divided images are joined and combined in the joining and combining order (step 63) to obtain original image data (step 64).

For example, as shown in FIG. 7, if the dispersion value of the density change is large in the area along the edge of the joining and combining portion, the joining of the images tends to be conspicuous. Therefore, the junction synthesis is started from this portion. Further, as shown in FIG. 8, in a region where the dispersion value of the density change is small, even a slight deviation in the image portion is allowed. Further, when the variance value of the density change is large, it is possible to reduce the error in the black and white matching at the edge portions of the adjacent areas, and the influence on the subsequent junction synthesis is small.

Another embodiment of the present invention will be described with reference to FIGS. 9 and 10. FIG. 9 is a flowchart for performing a process of blurring the joining and combining area when joining and combining the divided images. FIGS. 10A to 10C are explanatory diagrams of the joining composite area obtained by performing the blurring process on the joining composite area. In FIG. 9, steps 21 to 29 and 60 are the same as those shown in FIGS. 2, 4 and 6, and therefore description thereof will be omitted. In the present exemplary embodiment, after the distortion correction of the divided images is performed in order to join and combine the divided images, the density determination circuit 123 in the image processing unit 12 (see FIG. 1) causes the density combining circuit 123 to combine and combine the divided images. The pixel of interest in is extracted, and the density of each pixel in the vicinity of the pixel of interest is determined (step 71). For example,
A total of nine pixels, namely, eight pixels surrounding the pixel of interest "X1" shown in FIG. 10B, that is, eight pixels surrounding the pixel "X1" and the pixel of interest "X1" are extracted and their densities are determined. .. For example, when the image shown in FIG. 10B is a character, the character portion is black and the others are white, and thus the pixels near the pixel of interest “X1” have 5 black pixels and 4 white pixels.
It is a pixel. Then, the density average value calculation circuit 127 in the image processing unit 12 calculates the average density of the junction combining area from the densities of the respective pixels (step 72). afterwards,
The divided images are joined and combined according to the averaged density (step 73). For example, the target pixel “X1” in FIG. 10B has X1 = 5/9 from the number of black pixels in its vicinity, and an image output signal having a density of 5/9 of black pixels is output. Further, since the target pixel “X2” shown in FIG. 10C has seven black pixels in its vicinity, the average density is calculated as X2 = 7/9 in the same manner as described above, and the average pixel density is 7 / An image output signal having a density of 9 is output. Then, the target pixel in all the combined images of the divided images is subjected to the same processing as described above, and the original image data of the combined image area is output. As a result, for example, as shown in FIG.
Since the joining composite area of the divided images is output as the average value of the densities in the vicinity of each pixel, the image becomes a blurred image without a step. Therefore, in the splicing / synthesizing area of the divided images, an image with high fidelity can be obtained as compared with the case of simply joining. Although the embodiment of the character image has been described, the image quality can be improved by performing similar blurring processing on an image having gradation.

Another embodiment of the present invention will be described with reference to FIG. FIG. 11 is a diagram illustrating an illumination device of an image reading device according to an embodiment of the present invention. In FIG.
The base 81 is an unspecified flat case such as an office desk. The imaging scanning unit 83 is symmetrically placed on one end of the center line 82 of the base 81, and the diffused illumination 84 is symmetrically placed on the other end of the center line 82. Then, the document image 86 is placed at the center of the image capturing scanning unit 83 and the diffused illumination 84. That is, it is on the line segment in the direction in which the center of the original image 86 is seen from the imaging scanning unit 83 and in the symmetrical direction about the perpendicular line 85 at the center of the original image 86. The diffused illumination 84 uses, for example, a point light source whose main optical axis is the original image 8.
It is provided with a diffuser plate which is directed to the center of the set position of 6 and is sufficiently diffused so as to avoid a problem due to specular reflection light.

As described above, the imaging scanning unit 83 and the diffused illumination 8
4 is arranged, the imaging scanning unit 83 and the imaging scanning unit 83
The sum of the distance a from the reading unit close to the distance a and the distance a ′ from the reading unit to the diffused illumination 84 is the distance b between the imaging scanning unit 83 and the reading unit far from the imaging scanning unit 83, and from the reading unit. It is almost the same as the sum with the distance b ′ to the diffused illumination 84. Therefore, when the imaging scanning unit 83 and the diffused illumination 84 are arranged as shown in FIG. 11, the imaging scanning unit 83 is arranged.
The illumination distribution viewed from above is uniform over the entire surface of the document image 86. In particular, in the process of joining and combining the divided images, it is not necessary to mutually adjust the characteristics in the gradation direction across a plurality of images, and the processing load is reduced.

[0043]

According to the present invention, it is checked whether or not the entire surface of the divided image has a uniform density, and the distortion correction process is not performed on the divided image having the uniform density on the entire surface of the divided image. As a result, the processing load on the image reading apparatus and the processing time can be reduced at the same time. Further, since the joining and combining process of the divided images is started from the central portion where a relatively large number of images exist, the amount of deviation is small and inconspicuous in the central portion of the image. Further, in the joining / combining process of the divided images, a portion having a large dispersion value of the density change in which the deviation of the joining portion is noticeable is preferentially joined, and a portion of the dispersion value of the density change in which the deviation of the joining portion is not noticeable is joined. The quality of the image has been improved. Further, even if there is a step at the boundary portion of the joining composite area of the divided image, the average value of the densities in the vicinity of each pixel in the area is calculated, and the image data is output at the average density. An image with high quality can be obtained without the step of the above portion being conspicuous.

[Brief description of drawings]

FIG. 1 is a block configuration diagram for explaining an embodiment of the present invention.

FIG. 2 shows a flowchart for explaining an embodiment of the present invention.

FIG. 3 is an explanatory diagram when joining and combining divided images.

FIG. 4 is a flowchart in the case where the joining and combining order of divided images is started from the vicinity of the center.

FIG. 5 is an explanatory diagram in which the joining synthesis order of divided images is started from the vicinity of the center.

FIG. 6 is a flowchart in the case where the joining and combining order of divided images is started in the descending order of the density distribution value.

FIG. 7 is a diagram illustrating before and after joining of divided images having high density.

FIG. 8 is a diagram illustrating before and after joining of divided images with low density.

FIG. 9 is a flowchart for performing a process of blurring a joint compositing region when jointing and composing divided images.

10 (a) to 10 (c) are explanatory diagrams of the joining composite area obtained by performing the blurring process on the joining composite area.

FIG. 11 is a diagram illustrating an illumination device of an image reading apparatus that is an embodiment of the present invention.

FIG. 12 shows an image reading apparatus when the optical system is moved and the solid-state image sensor is fixed.

FIG. 13 shows an image reading device when an optical system and a solid-state image sensor are moved.

FIG. 14 is a schematic diagram of an image reading apparatus in a conventional example.

FIG. 15 is a diagram illustrating a scanning unit of the image reading apparatus in FIG.

FIG. 16 is an explanatory diagram of a joining boundary line when joining divided images in a conventional example.

[Explanation of symbols]

11: Input instruction unit 13: Imaging scanning unit 111: Resolution designation unit 131: Drive processing circuit 112: Reading start designation unit 132: Image sensor 12: Image processing unit 133: Image formation Optical system 121 Control circuit 134 Drive mechanism 122 Image processing circuit 135 Control circuit 123 Density determination circuit 14 Original image 124 Density correction circuit 125 Distortion correction processing circuit 126 .Junction synthesis order determination circuit 127 .. Concentration average value calculation circuit

Claims (5)

[Claims] 1. An image forming optical unit capable of varying a magnification, an image pickup unit arranged in a predetermined positional relationship with respect to the image forming optical unit, and the image forming optical unit and the image pickup unit are document images. A scanning means capable of scanning in the X-axis and Y-axis directions, and capable of dividing the original image into a plurality of images;
An image reading apparatus comprising a joining and combining processing unit that obtains one original image data by joining and combining the divided and captured divided images, and calculates the density of the divided images so that the entire surface has a uniform density. And a distortion correction processing means for omitting the distortion correction processing of the divided image which is determined to have a uniform density on the entire surface by the density determination means. Image reading device. 2. The image reading apparatus according to claim 1, wherein the splicing and combining of the divided images are processed in an order from a central portion toward a peripheral portion. 3. The image reading apparatus according to claim 1, wherein the adjacent areas when the divided images are combined and combined are imaged in advance by an image pickup unit so as to overlap by several pixels. 4. The image reading apparatus according to claim 1, wherein the splicing and combining of the divided images are performed in the order of increasing dispersion value of density change along edges of adjacent small areas of the divided images. 5. The joining synthesis of the divided images is performed by, when joining the divided images, extracting pixels in the vicinity of a target pixel in an adjacent region, calculating an average density of pixels in the vicinity, and calculating the average. The image reading apparatus according to claim 1, wherein the density is used as an image output signal of the adjacent area with the density of the pixel of interest.