JPH09261448A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an output image of high definition by plurally setting the range of recognizing an image area to excellently erase a frame area with respect to an original with plural image areas. SOLUTION: The plural ranges of recognizing an image area are provided to obtain information of a frame area within respective ranges, plural frame erasing processing is executed by leaving the individual image area and image data removed of a frame area within each range is generated to output to a laser printer. When a micro image is duplex, a reading range is divided into right left two pieces and respectively one black frame is searched in right and left frame recognizing area to obtain two pieces of square information (C). Thus, as the frame erasing processing of two black frames is easily executed by leaving two image areas, the image is not lost with respect to a microfilm (B) imprinted with two images as well and the black fame is erased (D) without leaving a space between the right and left images black to obtain the high definition output image.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention reads a document image printed in a duplex form on, for example, a microfilm, generates image data from which a frame area around the image area is removed, and outputs the image data to a printer or the like. The present invention relates to an image processing device.

[0002]

2. Description of the Related Art An image processing apparatus for processing an image projected on a microfilm has a function of enlarging and projecting a micro image on a screen and outputting image data read by a photoelectric conversion element to a printer side. There is a reader / scanner device having a function of printing out on recording paper. The image printed on the microfilm has a frame area around the image area.

In general, the reader scanner device is provided with a function of removing the image data corresponding to the frame area of the microfilm from the image data and preventing the frame area from being printed out on paper as black solid. ing. The reason why such a function is added is not only because the black solid portion is unsightly but also because of the demand for suppressing the toner consumption amount in the printer.

The conventional frame recognition method includes, for example, Japanese Patent Publication No.
As shown in No. 65371, there is a method in which an image is searched from the periphery of the reading range to the inside and the outside of the rectangle is whitened. There is also a method in which four sides of the image are searched so that the outside of the image can be erased even if the image is oblique.

By the way, although the micro image is used in various business operations, for example, in order to respond to a check inquiry business, the front and back surfaces of a check are printed as a set on one image frame of a microfilm. ing. Such a form in which the front and the back of the subject are photographed on the microfilm so that the front and the back are arranged in two rows is called a duplex.

Even in the case of a microfilm on which a plurality of images are imprinted as a set like such a duplex, the same problem occurs that the frame area around the image area is printed out as black solid on the paper. Therefore, it is strongly requested to erase the frame area with high quality.

[0007]

However, in the reader / scanner device adopting the frame recognition method disclosed in the above publication, if there are two images within the range read by the photoelectric conversion element, a frame area between the images is generated. However, there is a problem that a black solid portion remains between the images when the printout is made.

Further, the reader / scanner apparatus adopting a method of searching for the upper, lower, left and right sides one by one and recognizing the outside of the area surrounded by the found sides as a frame area (black frame),
The following problems occur. That is, for example, when two images are arranged side by side, the side of either one of the left and right images is adopted as the upper and lower sides, but at this time, the side of the other image is extended. If it does not exist on the line, there is a problem that an image loss occurs in the other image.

The present invention has been made in order to solve the problems associated with the prior art described above, and it is possible to obtain a high-quality output image by erasing a frame area of a document having a plurality of image areas with high quality. An object of the present invention is to provide an image processing device capable of performing the above.

[0010]

In order to achieve the above object, the present invention comprises a reading means for reading a document having a frame area around the image area, and the image is read based on an image signal from the reading means. In an image processing device for recognizing a region and generating image data in which the frame region is removed, setting means for setting a plurality of ranges for recognizing the image region, and recognizing the image region within each range set by the setting means An image processing apparatus, comprising: an image area recognizing unit that performs the image area recognizing unit; and a generating unit that generates image data in which the frame areas in the respective ranges are removed based on an output from the image area recognizing unit. Is.

In the image processing apparatus having the above-mentioned structure, the following acts on a document having two image areas on the left and right sides, for example. A document that has a frame area around the two image areas is
The whole is read by the reading means. The reading range read by the reading unit is divided into left and right by the setting unit, and two ranges for recognizing the image area are set on the left and right. The image area recognizing unit recognizes the image area in the left side range to obtain the position information of the first frame area, and further recognizes the image area in the right side range to detect the second frame area. Location information is required. Based on the output from the image area recognizing means, the generating means generates image data from which the frame areas in the respective right and left areas have been removed, and outputs this image data to a laser printer or the like.

As described above, the range in which the image area is recognized is divided into a plurality of areas, and the information on the frame area is acquired within each range. Therefore, it is possible to easily erase a plurality of frames while leaving the individual image areas. In addition, even for a document having a plurality of image areas, the image area is not lost, and the frame area between the images is erased, so that an output image with high image quality can be obtained. Can

[0013]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a perspective view showing an external appearance of an image processing system incorporating a reader scanner device according to an embodiment of the present invention, FIG. 2 is a schematic configuration diagram of an optical path system of the reader scanner device, and FIG. 3 is a reader scanner device. 3 is a front view showing a state in which a duplex image is projected on the screen of FIG.

<< 1. Mechanism of Reader Scanner Device >> The illustrated reader scanner device 1 irradiates a microfilm set on a microfilm carrier 4 and projects the transmitted light on a screen 9 for viewing.
Alternatively, it is a device that generates image data by projecting the transmitted light onto the CCD line sensor 12 (corresponding to a reading unit), and projects a micro image as follows.

As shown in FIG. 2, after the light emitted from the light source 2 is condensed by the condenser lens 3, the microfilm loaded on the carrier 4 is illuminated from below. The transmitted light of the microfilm is guided to the scanning mirror 7 via the projection lens 5 and the image rotation prism 6.

When the power is turned on, the reader mode is set as an initial state. In this reader mode, the film-transmitted light reflected by the scanning mirror 7 is further reflected by the reflection mirror 8 and reaches the screen 9, which is the screen. 9 is enlarged and projected. For example, when the front and back surfaces of a check are photographed on a microfilm, the array of microimages projected on the microfilm is in the form of duplex. In the case of a micro film on which a micro image is imprinted in the form of a duplex, in the reader mode, two images are enlarged and projected side by side on the screen 9 in parallel as shown in FIG.

On the other hand, when a reading mode is instructed from an operation switch or the like provided on an operation panel (not shown),
The scanning mirror 7 is rotated, and the film transmitted light reflected by the rotating scanning mirror 7 is further reflected by the plurality of reflecting mirrors 1.
It is reflected by 0 and 11 and guided to the CCD line sensor 12, and an image is formed on the sensor 12. As described above, in the reading mode, the image on the microfilm is scanned as the scanning mirror 7 rotates, and the image on the microfilm is scanned by CC.
Projection exposure is performed on the D-line sensor 12 and the lines are read one by one.

The image is scanned by the CCD line sensor 12 in the main scanning direction. Further, as the scanning mirror 7 rotates, the image is scanned in the sub-scanning direction perpendicular to the main scanning direction. The sub-scanning may be performed by driving the CCD line sensor 12 or the microfilm in addition to the mirror scanning for driving the scanning mirror 7.

In the CCD line sensor 12, 4800 photoelectric conversion elements are linearly arranged, and each of the elements converts the irradiation light amount into a voltage and outputs it. The light transmitted through the film forms an image on the CCD line sensor 12, and the brightness of one line on the film is converted into 4800 voltages, whereby the image of the microfilm is captured by the sensor 12 as an analog electric signal. Here, it is assumed that the output voltage of the sensor 12 is low when it is bright and high when it is dark.

FIG. 4 is a schematic block diagram of a control system of the reader / scanner device.

The A / D converter 30 converts the voltage output from the CCD line sensor 12 into an 8-bit digital image signal based on the conversion signal from the CPU 31. A
The / D converter 30 equally divides a certain upper limit voltage value and a lower limit voltage value into 256 pieces, and outputs a predetermined digital value based on which section the input analog voltage corresponds to. Here, the one with a low output voltage (bright one) from the CCD line sensor 12 is 0, and the one with a high output voltage (dark one) is 255.
And

The converted digital image signal is input to the first processing circuit 32 and the second processing circuit 33, and the CPU 31
A predetermined image processing such as gamma conversion, enlargement / reduction, median processing, edge enhancement, frame elimination, and binarization is performed on the basis of the signal output from, and then output to the outside. Although the order of each image processing is appropriately determined, the image processing required for sampling in the preliminary scan is performed by the first processing circuit 32, and the other image processing is performed by the second processing circuit 32.

The digital image signal at the time of preliminary scanning is subjected to predetermined image processing by the first processing circuit 32, then sampled at a predetermined pitch and stored in the image memory section 34. The CPU 31 calculates the sampling data read from the image memory unit 34 based on a program stored in the EPROM 35, and performs a frame erasing process for erasing a black frame. The CPU 31 outputs a control signal to a motor that drives the scanning mirror 7, a motor that drives the image rotation prism 6, and the like, while the CPU 31 receives a fixed position signal of the scanning mirror 7 and the prism 6. In addition, EP
The ROM 35 stores a program for performing a frame erasing process described later, and the RAM 36 is a C
This is for temporarily storing the calculation result and the like when the PU 31 performs various processes.

The image data subjected to the predetermined image processing in the second processing circuit 33 is output through the interface circuit and transferred to the laser printer 20 shown in FIG. 1 to perform copying or the like based on the image data. , Transferred to a personal computer and displayed on a CRT, or stored in an optical disk storage device. The image data may be temporarily stored in the RAM in the interface circuit.

The CPU 31 removes the setting means for setting a plurality of ranges for recognizing the image area, the image area recognizing means for recognizing the image area within each set range, and the frame area within each range. It functions as a generation unit that generates image data.

<< 2. Outline of Processing >> Next, an outline of processing in the reader scanner device will be described based on the flowchart shown in FIG.

In the reader / scanner apparatus 1 of the present embodiment, when a microfilm printout is instructed by an operation switch (not shown) or the like, first, a preliminary scan is performed, and after a predetermined process, a main scan is subsequently performed. Done.

First, it is determined whether the target microfilm is a negative film or a positive film (S11).

Next, frame recognition is performed (S12). This frame recognition processing is performed depending on whether the micro image array is a simplex film or a duplex film. In the case of simplex film,
The entire reading range of the CCD line sensor 12 is selected as a frame recognition area, a black frame is detected in this frame recognition area, and frame recognition processing is performed to obtain position information of the black frame. On the other hand, in the case of a duplex film, first, the entire reading range is divided into left and right parts, that is, a left frame recognition region and a right frame recognition region. First, the left-side frame recognition area is designated or selected, a black frame is detected in the left-side frame recognition area, and a frame recognition process for obtaining position information of the first black frame is performed. Next, the right frame recognition area is switched to, and a black frame is detected in the right frame recognition area,
A frame recognition process for obtaining position information of the th black frame is performed.

When the frame recognition (S12) is completed, next, A
An E (automatic exposure) process (S13) is performed to set the light amount of the light source 2 during the main scan. This setting is performed based on only the data in the image area identified by the frame recognition process (S12). This enables optimum exposure control that is not affected by the frame.

After that, centering for printing on the central portion of the paper at the same size, or full framer for printing by enlarging the size of the output paper, depending on the image output range set according to the mode. To print, the main scan for the printing operation is performed (Sl
4).

The image data read by the main scan is subjected to predetermined image processing, and further, step S12.
A frame erasing process is performed according to the number of images based on the frame recognition information obtained in step. In other words, in the case of simplex film, one frame erase process is performed based on the frame recognition information,
In the case of a duplex film, two frame erasing processes are performed based on the frame recognition information obtained in each of the left and right frame recognition areas, and then output to the laser printer 20.

<< 3. Details of Processing >><< 3-1. Negative / Positive Discrimination (S11) >> When the negative / positive is designated by an operation switch (not shown), the negative / positive discrimination in step S11 is performed based on the designation. Be seen. If there is no designation, negative / positive discrimination is performed by comparing the lightness and darkness of the entire microfilm with a threshold value during prescanning (preliminary scanning).

It is necessary to first perform the negative / positive determination in the frame recognition process (S12) executed subsequently, depending on whether the film is a negative film or a positive film. This is because the content changes.

That is, as shown in FIG. 7, in the negative film, the frame is the low density portion and the image area is the high density portion (hatched portion), while in the positive film, the frame is the high density portion and the image area is opposite. Is the low concentration part. Then, in the frame recognition process (S12), in the case of a negative film, “frame →
Data (digital value) is "small (bright)" at the boundary of "image area"
→ While searching for a frame by utilizing the change of "large (dark)",
In the case of a positive film, the frame is searched by utilizing the fact that the data changes from “large (dark) to small (bright)” at the boundary of “frame → image area”. Since the processing contents of frame recognition are changed in this way, it is necessary to first discriminate negative / positive.

<< 3-2. Frame Recognition (S12) >> FIG. 16 is a flowchart showing the procedure of frame recognition processing, and FIG. 17 is shown in FIG.
6 is a flowchart showing a procedure of image area extraction processing shown in FIG.

In the frame recognition in step S12, whether the frame recognition region is the entire reading range or the left and right divided regions is preset before the image region extraction processing.
Based on this setting, the image area extraction process for identifying the frame and the image area is executed. Whether the determination form of the frame recognition area is the entire reading range or a divided range is selected or designated by an operator operating an operation switch or the like (not shown). The image area extraction process is shown in FIG.
It will be described later based on 7.

As shown in FIG.
In the case of a film, two images are projected on the left and right on the screen 9. When the duplex mode is designated by the operation switch or the like (S21; YES), first, the left half (ABED of FIG. 3) of the entire reading range is set as the frame recognition area (S22), and the left frame recognition area is set. An image area extraction process is performed on the position information to obtain position information of the first black frame (S23). When the image area extraction processing for the left frame recognition area is completed (S22, S2
3) Next, the right half of the entire reading range (see B in FIG. 3).
CFE) is set as the frame recognition area (S24), and the image area extraction processing is performed on this right frame recognition area, and 2
The position information of the th black frame is obtained (S25).

On the other hand, when the duplex mode is not designated (S21; NO), the entire reading range (ACFD in FIG. 3) is set as the frame recognition area (S2).
6), the image area extraction processing is performed on this frame recognition area, and the position information of one black frame is obtained (S27).

The left and right frame recognition areas to be divided are not limited to the size of 1/2 of the reading range, but may be set to a slightly wider range.

Image region extraction processing (S23, S25, S2
In 7), as shown in FIG. 17, first, candidates of boundary lines between the frame and the image area, that is, coordinate groups estimated to constitute the upper, lower, left, and right boundary lines are detected from the upper, lower, left, and right sides, respectively. (S31). Next, it is checked whether or not the detected coordinate groups are continuous, and if there are coordinates at positions far apart in the continuous coordinate group, the coordinates are corrected to positions on the continuous line (S3).
2). Next, linear equations representing the upper, lower, left, and right boundaries are calculated (S33). This calculation is performed based on the continuous coordinate group having the largest number of coordinates among the continuous coordinates. Finally, the coordinates of the intersections (four corners) of the upper, lower, left, and right boundaries are calculated (S34), and the frame and the image area are identified.

The advantage of designating the frame recognition area according to whether or not the micro image is duplex as in the present embodiment will be described. As conceptually shown in FIG. 5A, when there is only one image in the reading range, the sides indicated by the symbols, and are the left side, the lower side,
Since they are recognized as the right side and the upper side, respectively, the frame erasing process can be favorably performed, and the output image as shown on the right side can be obtained.

On the other hand, as conceptually shown in FIG. 5B, in the case where the left image and the right image are lined up in the X direction in the reading range and the frame recognition region is not designated to be divided, The following problems occur. In this figure, the sides indicated by the symbols, and are the left side, the bottom side, the right side, and the top side of the left image, and the sides indicated by the symbols, and are the left side, the bottom side, the right side, and the top side of the right image. And

When the image area extraction processing is performed without designating the frame recognition area as divided, the left side can be recognized as the left side (side) of the left image and the right side can be recognized as the right side (side) of the right image. When checking the continuity of candidate points, the lower side is recognized as the lower side (side) of the left image and the lower side (side) of the right image. The longer side is used as the lower side in this frame recognition area. To do. Similarly, the upper side (or) of the left or right image is adopted as the upper side. If the upper and lower sides of the left image are longer than the upper and lower sides of the right image, then the side is adopted, and as a result,
Recognized as a rectangle made with and. At this time, the output image is as shown on the right side, the upper part of the right image is lost, and the black frame remains below. Furthermore, there is a problem that the left and right images also remain black.

On the other hand, in the present embodiment, as conceptually shown in FIG. 5 (C), when the micro image is a duplex, the reading range is divided into two left and right, and the left and right are read. By searching for one black frame in each frame recognition area, two pieces of quadrilateral information are acquired. For this reason, it is possible to easily perform the frame erasing process of the two black frames by leaving the two image areas. Therefore, the same can be applied to the microfilm on which the two images shown in FIG. As shown in FIG. 3D, the image is not lost, and the black frame is erased without left black between the left and right images, so that a high-quality output image can be obtained.

In the present embodiment, the reading range (see FIG.
In addition, in order to search for two frames in FIG. 5 (ACFD), the frame recognition area is divided into two left and right at the center of the reading range, and one of the left area (ABED) and the right area (BCFE) is divided. The case where the frame is searched has been described. Here, the frame recognition area is divided into two on the premise that it is two images, but for a plurality of images of three or more, the reading range is further divided according to the number of the images. Increase the recognition area.

Further, the form of discriminating the frame recognition area is not limited to the designation by the operation switch or the like. For example, by examining the sampling data, it is determined whether or not the image is divided into one, and if the result of this determination is that there is one image, the frame recognition area remains the reading area (ACFD),
Subsequent processing is automatically performed. If there are a plurality of images, it is possible to divide the reading range according to the number and arrangement state of the images, set a plurality of frame recognition areas, and automatically perform the subsequent processing.

<< 3-3. Image area extraction (S23, S25,
S27) >> The details of the image area extraction processing (FIG. 17) will be described below with reference to step S31, step S32, step S3.
3 and step S34 will be described in this order.

<< 3-3-1. Searching for Candidates (S31) >> The image signal prescanned from the lower end side of the image and read by the sensor 12 is processed as described above to be image data, Like No. 6, it is fetched at a pitch of 1 mm and stored in the edit RAM. For example, in the case of 400 dpi, since one dot is 63.5 μm, it is 15.748.
It is captured for each dot. As shown in FIG. 8, this edit RAM can store image data from the address (00,00) to the address (210,297). this is,
A4 size is 210 x 297 mm, so RAM required when storing at a 1 mm pitch as described above
Capacity.

A coordinate group estimated to form a boundary line between the frame and the image area (upper and lower left and right boundary lines), that is, a frame candidate coordinate group is detected based on the data in the edit RAM. For example, in the case of detecting the frame on the left side of the image area as shown in FIG. 18, as shown by the broken line arrow group A in FIG. Data is searched for in the direction. In the duplex mode, when the left and right frame recognition areas are divided into two reading areas, the reading is performed by dividing the lines into 0 to 105 lines and 106 to 210 lines in the X direction.

First, the coordinates of the left end of the line (X = 0
0) is set as the reference coordinate, and the data is read (S102). Further, the data of the second coordinate (X = 02 in this case) from the reference coordinate in the X direction is read (S103). Next, the two read data are compared (S104). As a result, if the difference is less than or equal to a predetermined value (= reference value) (S105; NO), the reference coordinate is moved one coordinate in the X direction (S109; in this case, X = 0.
1) After the data of the coordinates are read (S10
2) Similarly to the above, the second coordinate in the X direction (in this case,
It is compared with the data of X = 03 (S103, S10).
4).

In this way, the reference coordinates are moved one by one in the X direction, and the comparisons are sequentially performed. As a result, when the difference between the two data read out first exceeds the reference value in the line (S105; YES), a position between the two coordinates is a frame candidate on the left side of the line. The coordinates are stored and stored (S106).

Next, the object to be processed is moved by one line in the Y direction (S107), and the line after the movement is
Similar to the above, the comparison is sequentially performed from the left end side (S102).
~ S104), frame candidate coordinates on the left side of the line after the movement are searched (S105). Although not shown in the processing of FIG. 18, even if the reference coordinates become the second coordinates from the right end as a result of the reference coordinates being moved in the X direction by searching as described above, When no frame candidate coordinate on the left side is detected on the line, the processing target is naturally moved by one line in the Y direction.

In this way, all lines in the Y direction (0th line
When the search for the left-side frame candidates for lines (line 297) is completed (S08; YES), the right-side frame candidate coordinates are next, then the upper-side frame candidate coordinates, and finally the lower-side frame candidate coordinates. , Searched.

The search for the frame candidate on the right side is performed by searching the data in the negative direction of X for each line in the Y direction as shown by the group of broken line arrows B (FIG. 9). Each line in the X direction like the group (0th line to 210th line)
The search for the frame candidate on the lower side is performed by searching for data in the Y direction with respect to each line) by searching for the data in the negative direction of Y for each line in the X direction as indicated by the group of broken line arrows D.

In this frame candidate search processing (S31), the frame candidate coordinates are not searched by comparing the data of the reference coordinates with the data of the next coordinates, but as described above, skip the next (skip one). I am looking for it by comparing it with the data of the second) coordinate. This is because the density change at the boundary between the frame and the image area is comparatively gentle, and thus the difference of the predetermined value (= reference value) cannot be obtained between the data of the adjacent coordinates.

<< 3-3-2. Check if continuous (S3
2) >> With respect to the frame candidate coordinates searched for in step S31, continuity is sequentially checked by the procedure shown in the flowcharts of FIGS. 19 and 20. Hereinafter, description will be made with reference to FIG.

First, the first data is used as a reference, and its coordinate value is compared with the coordinate value of the next data (S20).
3). As a result, if the difference between the X-direction and the Y-direction is 1 or less (1 mm or less), these two data are considered to be continuous (S204; YES) and saved (S24).
1) In addition, the continuous number is counted (S242). Further, the last data stored (in this case, the next data described above) is used as a reference (S243), and the comparison is performed in the same manner as above. That is, the coordinate value of the newly set data is compared with the coordinate value of the next data (S20
3), processed in the same manner. For example,
The case is discontinuous in the X direction but continuous in the Y direction. Further, ˜ is a case where the X direction and the Y direction are continuous.

If the result of comparison is that the difference exceeds 1, (S204; NO), one is skipped and compared with the coordinate value of the next data (S205). As a result, if the difference is 2 or less (2 mm or less), it is determined that these data are continuous (S206; YES), and the data skipped above is the middle of the two data to be compared. The coordinates are corrected to (S221). After that, the data is stored (S241), and after the continuous number is counted (S24).
2) The last stored data is used as a reference (S24
3) Based on this newly set reference data, the same processing as above is continued. For example, ~ in Fig. 10 is discontinuous, and l0 surrounded by ~ ○ is also discontinuous, but since the difference between and 10 surrounded by ○ is 2 in the Y direction, it is regarded as continuous ( (S206; YES), the coordinates are corrected to the intermediate coordinates of 10 surrounded by G and O (S221).

If it is determined in step S206 that the data is discontinuous (S206; NO), another one is skipped and the comparison with the coordinate value of the next data is performed (S2).
07). As a result, if two intermediate data are skipped and compared, it is determined that they are continuous (S208; YES),
The two data in between are corrected in the same manner as above (S
231), then data storage (S241), continuous count (S242), and reference data update (S24).
3) is performed as described above. In this way, the coordinates of one or two intermediate data are corrected (S22
1, S231)), the adverse effect when the frame candidate coordinates are erroneously detected due to scratches or dust on the microfilm is prevented.

On the other hand, when it is determined that the data is discontinuous also in step S208 (S208; NO), it is determined that the data up to the reference is continuous, and the next data is not continuous (S209). . Therefore, the storage location of the continuous coordinate group is moved (S210), the data next to the reference data is set as the reference, and a new continuous coordinate group is newly set based on the newly set reference data. Is searched for (S211).

<< 3-3-3. Calculating linear equation (S33) >>
The continuous coordinate group having the largest number of continuous ones is picked up from the plurality of continuous coordinate groups searched in step S32, and the frame and the image area are processed based on the continuous coordinate group by the procedure shown in the flowchart of FIG. A straight line expression representing the boundary of is calculated for each of the upper, lower, left and right sides. However, if the number of consecutive picked-up continuous coordinate groups is less than a certain reference number, no calculation is performed on the sides of the consecutive coordinate group. This is because it is uncertain whether or not it is a true frame when the number of consecutive characters is too small. Hereinafter, description will be made with reference to FIG. 11 as well.

First, the maximum value, the minimum value, the difference and the average value of the coordinates are calculated for the picked-up continuous coordinate group (S301). Here, when the linear expressions of the left side and the right side are calculated, the X coordinate is calculated, and when the linear expressions of the upper side and the lower side are calculated, the Y coordinate is calculated.

Next, it is judged whether or not the difference between the maximum value and the minimum value is 2 or less, and if it is 2 or less (S302; YE).
S), the linear equation is not calculated, and the coordinates of the average value of the side are
The section of the side is shown (S307). That is,
The side is treated as being not inclined with respect to the coordinates of the edit RAM. This is because, due to the configuration of the sensor and photometry, even if there is no inclination, a variation of about 1 coordinate will always occur. Therefore, if the difference is 2 or less, it is treated as if there is no inclination. It is designed to prevent black frames and defects.

On the other hand, if the difference exceeds 2, (S30
2; NO), the X / Y coordinates of the fifth data from the beginning of the continuous coordinate group and (S304), the X / Y coordinates of the fifth data from the end (S305), and the side corresponding to the continuous coordinate group. Is calculated (S306). That is, the linear equation is not calculated based on the coordinates of both ends of the continuous coordinate group, but the linear equation is calculated based on the coordinates slightly inside. By doing so, it is possible to avoid the influence of the broken ear and the influence of the shift of the image edge at the time of photometry.

<< 3-3-4. Find the coordinates of the intersection (S3
4) >> The coordinates of the four corners of the image area are calculated by the procedure shown in the flowcharts of FIGS. 22 to 24 based on the linear equation of each side calculated in step S33. Hereinafter, FIG.
Description will be made also with reference to FIG.

When all the straight line equations indicating the upper, lower, left and right sides have been obtained (S40l; YES), the coordinates of the four corners are calculated from their intersections (S402).

If only three sides of the upper, lower, left and right sides are obtained (S403; YES), the coordinates of two points are obtained as the intersections of the linear equations of the three sides (S411). Another 2
The point can be obtained from the frame candidate coordinate group as the coordinate of the end on the opposite side of the two points obtained as the intersection (S412).

For example, the upper two points in FIG. 12 are respectively determined as the intersection of the left side and the upper side and the intersection of the right side and the upper side. Further, the other two points are respectively obtained as the coordinates of the lower end of the left side and the coordinates of the lower end of the right side. Here, the coordinate of the end (lower end) on the opposite side is detected as a frame candidate coordinate of the relevant side in the process of searching for a candidate (S31), and if necessary, in the process of checking whether it is continuous (S32). Figure 1
In the coordinate group corrected as 0, it is the coordinate at the opposite end. The accuracy can be increased by checking whether the left side and the right side are parallel to each other.

If only two sides are obtained (S40
5; YES), different processing is performed depending on whether the two sides are parallel or two orthogonal. In the case of two parallel sides (S421; YES), the coordinates of both ends thereof are obtained from the frame candidate coordinate group as shown in FIG. 14 (S4).
22). Here, the coordinates of both ends are the same as the definition of "end" in "when three sides are determined". Two orthogonal
In the case of a side (S421; NO), as shown in FIG. 13, the coordinate of one point is obtained as the intersection of the linear equations of the two sides (S42).
3). Further, two points adjacent to this one point are respectively obtained from the frame candidate coordinate group as coordinates of the ends on the opposite side of the two sides (S424). The remaining one point is calculated based on the above three points (S425).

As described above, even when not all the sides are obtained, the four corners of the image area can be determined when the two sides or the three sides are obtained. The frame can be recognized even in the image.

It should be noted that only one side is obtained, or
If no side has been obtained (S405; NO), it is determined that there is no frame (S407). In this way, in the present embodiment, the boundary line between the frame and the image area is determined.

[0073]

As described above, according to the image processing apparatus of the present invention, the range in which the image area is recognized is divided into a plurality of areas, and the information on the frame area is acquired within each range. It is possible to easily and with high quality perform a plurality of frame erasing processes while leaving the image area of. Therefore, for a document having a plurality of image areas, there is no loss of the image area,
Moreover, it is possible to obtain an output image with high image quality in which the frame area between the images is erased.

[Brief description of drawings]

FIG. 1 is a perspective view showing an external appearance of an image processing system incorporating a reader / scanner device according to an embodiment of the present invention.

FIG. 2 is a schematic configuration diagram of an optical path system of a reader scanner device.

FIG. 3 is a front view showing a state in which a duplex image is projected on the screen of the reader scanner device.

FIG. 4 is a schematic block diagram of a control system of the reader scanner device.

5A to 5D are explanatory diagrams for explaining the advantage of designating a frame recognition region depending on whether a micro image is a duplex or not.

FIG. 6 is an explanatory diagram showing a data reading pitch.

FIG. 7 is an explanatory diagram showing a state in which the density of a frame is reversed on a negative / positive film.

FIG. 8 is an explanatory diagram showing coordinates on the edit RAM.

FIG. 9 is an explanatory diagram showing that when a coordinate forming a boundary line between a frame and an image area is searched, the left side is searched from the left side, the right side is searched from the right side, the upper side is searched from above, and the lower side is searched from below. .

FIG. 10 is an explanatory diagram of a method of determining whether or not the coordinate groups forming the boundary line are continuous.

FIG. 11 is an explanatory diagram of a method in which the inclination is 0 ° or 90 ° when the difference between the maximum value and the minimum value of the X or Y coordinates of the coordinate group forming one side of the boundary line is 2 or less. .

FIG. 12 is an explanatory diagram showing a method for obtaining coordinates of four corners of an image area when three sides of a boundary line are found.

FIG. 13 is an explanatory diagram showing a method for obtaining coordinates of four corners of an image area when two sides of a boundary line are found.

FIG. 14 is an explanatory diagram showing a method of obtaining coordinates of four corners of an image area when two sides of a boundary line are found.

FIG. 15 is a flowchart showing an outline of an operation sequence of the reader scanner device.

FIG. 16 is a flowchart showing a procedure of frame recognition processing.

FIG. 17 is a flowchart showing a procedure of image area extraction processing.

FIG. 18 is a flowchart showing, on the left side, a procedure for searching for coordinates that form a boundary line between a frame and an image area.

FIG. 19 is a flowchart showing a procedure for determining whether or not the coordinate groups forming the boundary line are continuous.

20 is a partial diagram of FIG.

FIG. 21 is a flowchart showing a procedure for calculating an expression representing a boundary line.

FIG. 22 is a flowchart showing a procedure for obtaining coordinates of four corners of an image area.

FIG. 23 is a flowchart showing a procedure for obtaining coordinates of four corners of an image area when three sides of a boundary line are found.

FIG. 24 is a flowchart showing a procedure for obtaining coordinates of four corners of an image area when two sides of a boundary line are found.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Reader scanner device (image processing device) 2 ... Light source 7 ... Scanning mirror 9 ... Screen 12 ... CCD line sensor (reading means) 31 ... CPU (setting means, image area recognition means, generation means)

Claims (1)

[Claims] 1. A reading means for reading a document having a frame area around the image area is provided, and based on an image signal from the reading means, the image area is recognized and image data in which the frame area is removed is generated. In the image processing apparatus, setting means for setting a plurality of ranges for recognizing image areas, image area recognizing means for recognizing image areas within the respective ranges set by the setting means, and output from the image area recognizing means And a generating unit that generates image data in which the frame area in each of the ranges is removed based on the above.