JPH08107479A - Image processing unit - Google Patents

Abstract

PURPOSE: To generate a joint image without high quality without deviation in connection by correcting a position deviation even when an error is included in a detected scanning position and preventing the occurrence of a map missing region caused by the correction of the position deviation. CONSTITUTION: A position deviation detection circuit 7 detects a position deviation based on image data 400 in a scanning region in which a line image sensor 1 reads data in duplicate and image information of data stored in an image memory 6. Then the position deviation detection circuit 7 corrects a scanning position coordinate 300 to cancel the position deviation and provides the output of a corrected coordinate 710 to a mapping circuit 5. The mapping circuit 5 stores the image data 400 according to the corrected coordinate 710 to the image memory 6.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for storing image data read by scanning an original image in an image memory based on a scanning position corresponding to the image data.

[0002]

2. Description of the Related Art Conventionally, in a hand scanner that scans a document manually to read a document image, a scanning position of a reading sensor on the document is sequentially detected, and image data is imaged based on the detected scanning position. Was stored in memory.

As a method for detecting the scanning position of the reading sensor on the original, a sheet on which a reference grid is printed is used as an auxiliary device, and a reading sensor for detecting the reference grid on the sheet placed on the original is used. There is a method of calculating the position of the image data with respect to the reference grating from the output of the photosensitive element by scanning with the integrated photosensitive element (for example, US Pat. No. 4,260,979).

Further, a tablet is used as an auxiliary device,
By scanning the document placed on the tablet with the detection coil that is integrated with the reading sensor, the coordinate position on the tablet is detected from the output of the detection coil, and the image is displayed based on the detected coordinate position. There are methods of calculating the location of the data (eg, US Pat. No. 4,58).
No. 1, 761) Also, the position of the reading sensor is detected from the output of the encoder that is attached to the wheel that moves integrally with the reading sensor and that generates a pulse in accordance with the rotation of the wheel, and the image data is based on the detected position. There is a method for calculating the position of (for example, Japanese Patent Laid-Open No. 62-15964).

If an image processing apparatus that calculates the coordinate position of image data based on the scanning position detected by these conventional detection methods and stores it in a predetermined address of the image memory is used, a camera shake or meandering caused by manual scanning is used. Since it is possible to eliminate the influence of, the image distortion does not occur in the data stored in the image memory.

[0006]

However, the above-mentioned conventional structure has the following problems. When using an auxiliary device such as a sheet or tablet on which the reference grid is printed, precise processing and an expensive correction circuit are required to improve the scanning position detection accuracy and reduce the absolute position error.
Usually the cost is high. In addition, if an auxiliary device with low accuracy is used, a large absolute position error occurs in the detected scanning position.

On the other hand, when an encoder mounted on a wheel that moves integrally with the reading sensor and generating a pulse as the wheel rotates is used, the pulses of the encoder that are sequentially detected are counted with the scanning start position as a reference point. Since the scanning position is calculated in this manner, a large cumulative position error occurs due to mechanical accuracy, wheel slippage, and the like.

When the absolute position error or the cumulative position error of the detected scanning position becomes large, a large position error occurs between the scanning sensor and the regular scanning position on the original which is actually scanned by the reading sensor (hereinafter referred to as this Positional error is called "positional shift".)

When the coordinate position of the image data is calculated based on the misaligned scanning position and stored in a predetermined address of the image memory, a large image distortion occurs in the plane image reproduced on the image memory. There is. FIG. 19 shows the image distortion of the plane image due to this positional deviation.
An explanation will be given using (b) to (d).

A case where the hand scanner 1004 freely scans the original 1000 as shown in FIG. 19B will be described. As shown in FIG. 19C, in the case of a small hand scanner that reads in only one direction, the original 1
It is not possible to read the large screen above 000. However, since the image data is stored in the image memory in the reading range of the sensor in one direction, the connection shift does not occur. As shown in FIG. 19D, in a hand scanner that reads a large screen on the original 1000 by reciprocal scanning with a single stroke, when image data is stored in an image memory based on a scanning position that is displaced, an image connecting portion is connected. The image is distorted due to misalignment. In particular, in a scanner that generates a cumulative position error, the wider the scanning range, the more the position shift becomes, and the image itself cannot be read.

Next, problems in storing image data in the image memory will be described. As shown in FIG. 22,
A location 2002 where image data is stored in the image memory 2001 (image data storage pixel) and a location where image data is not stored (hereinafter referred to as “mapping area”) 200.
If there is 3, white streaks are noticeable in a solid black area, and image quality is significantly deteriorated. Further, as shown in FIG. 23, when the mapping missing area occurs in the character area, the quality of the character is remarkably deteriorated such as white streaks in the character.

In order to solve the above-mentioned problems in the prior art, the present invention corrects the positional deviation even if the detected scanning position includes an error, and the occurrence of a mapping missing area caused by the positional deviation correction. It is an object of the present invention to provide an image processing apparatus capable of generating a high-quality connection image without connection deviation by preventing the above.

[0013]

In order to achieve the above object, the structure of an image processing apparatus according to the present invention includes an image data read by scanning an original image and a scanning position corresponding to the image data. An image processing apparatus for sequentially inputting and storing the image data in an image memory based on a scanning position, wherein the image data and the storage data stored in the image memory in an overlapping scanning area to be scanned in duplicate. Position deviation detecting means for detecting the position deviation of the scanning position from the output and outputting a correction value for correcting the position deviation, correction means for correcting the scanning position based on the correction value, and the corrected scanning position. And a mapping means for storing the image data in the image memory based on the image data.

In the configuration of the device of the present invention, correction control means for varying the correction value at a predetermined ratio is further provided.
It is preferable that the correction unit corrects the scanning position based on the correction value that is changed at a predetermined ratio. Further, in this case, the correction control means uses the value obtained by dividing the correction value from the positional deviation detection means as the variable control amount, and changes the correction value stepwise at a predetermined rate for each different scanning position. Is preferred.

Further, in the configuration of the device of the present invention,
Delay means for delaying the corrected scanning position and the corresponding image data is further provided, and the mapping means stores the corresponding image data in the image memory on the basis of the delayed scanning position, thereby the position of the scanning position. It is preferable that the position for detecting the deviation and the storage position for storing the image data in the image memory are different.

Further, in the configuration of the device of the present invention,
The misregistration detection unit detects the scanning confirmation information of the scanning confirmation information memory and the scanning region holding unit that sequentially stores the scanning confirmation information of the image data input to the image memory in the scanning confirmation information memory, thereby overlapping the scanning confirmation information. It is preferable to include an overlap area detection unit that detects an overlap scan area that is scanned. Further, in this case, it is preferable to share the address of the image data stored in the image memory and the address of the scanning confirmation information stored in the scanning confirmation information memory. Further, in this case, there is further provided memory control means for controlling the storage of the image data in the image memory according to the determination of the scanning area performed by the overlapping area detection means, and the memory control means is configured to perform overlapping scanning in which overlapping scanning is performed. It is preferable to prohibit the storage of the image data in the image memory when it is determined that the area is the area.

Further, in the structure of the device of the present invention,
The misregistration detection means sets the storage address of the image memory corresponding to the coordinate value on the document of each pixel in the image data calculated based on the scanning position as a pixel position of interest, and its peripheral pixel positions including the pixel position of interest. On the other hand, an image correlation unit that detects a correlation value between the image data and the storage data stored in the image memory, and a correction amount calculation unit that generates a preset correction value from a plurality of the correlation values. And preferably. Also, in this case,
It is preferable that the misregistration detection unit generates a plurality of correlation values only in the case where the pixel positions to be correlated are all peripheral pixel positions including the pixel position of interest in the overlapping scanning region.

Further, in the structure of the device of the present invention,
It is preferable that the correction unit includes a holding unit that generates a cumulative correction value by accumulating the correction values output from the positional deviation detection unit, and corrects the scanning position by the cumulative correction value.

Further, in the configuration of the device of the present invention,
The mapping means includes an address generating means for generating a storage address of an image memory corresponding to a coordinate value of each pixel in the image data on the document based on the scanning position, and each pixel in the image data of the image memory. It is preferable to include a data storage means for storing at the storage address. Also,
In this case, the mapping means includes a pixel density conversion means for densifying the image data from adjacent pixels by interpolation, and a real number on the original of each pixel in the densified image data based on the scanning position. A unit for calculating coordinate values in units of the predetermined pixel density; an integer unit for converting real number coordinate values on the original document into integer coordinate values in units of the predetermined pixel density; A means for calculating a coordinate error value generated at the time of conversion into an integer, a means for converting the integer coordinate value into an address of an image memory, and a means for comparing the magnitude of the coordinate error value with a predetermined value. It is preferable that only the pixels whose magnitude of the coordinate error value is equal to or smaller than the predetermined value are stored in the address of the image memory. In this case, the pixel density conversion means further includes means for generating an interpolated pixel by using the pixel to be processed and the adjacent pixels around the pixel to be processed, and the image data is converted to a double pixel density by the means. Preferably, further,
The pixel to be processed P _{i, j} (however, in the image data, the j-th pixel data of the image data of the i-th line is shown) and its neighboring pixels P _{i, j + 1} , P _{i-1, j} , Using P _{i- 1, j + 1} ,
It is preferable to calculate the interpolated pixels Q _{i, j} , R _{i, j} and S _{i, j} from the above (Equation 1). Further, in this case, the predetermined value is 0.3 to 0.3 in units of the predetermined pixel density.
It is preferably 5.

Further, in the configuration of the device of the present invention,
The image data and the scanning position are preferably input from an input device equipped with an image sensor for reading an image and position detecting means for detecting the reading position of the image sensor by two detecting means. Further, in this case, the image sensor is preferably a line sensor. Further, in this case, the input device is preferably a hand scanner. Further, in this case, it is preferable that the detecting means is composed of a wheel that comes into contact with the document surface and rotates, and an encoder that outputs a pulse corresponding to the rotation of the wheel. In this case, further, it is preferable that the encoder outputs two types of pulses having different phases. Also,
In this case, it is preferable that the input device is detachably attached to the main body of the portable information device. In this case, further, by inserting the lock pin provided in the input device into the pin insertion hole provided in the body of the portable information device,
The input device is preferably attached to the main body of the portable information device.

In addition, in the configuration of the device of the present invention,
Coordinate rotation conversion means for rotating the scanning position at a plurality of predetermined angles about a predetermined coordinate is further provided, and the position shift detection means is configured to detect the image based on the coordinates of the scanning position corresponding to the plurality of predetermined angles. The coordinate value of each pixel in the data on the original is calculated, the storage address of the image memory corresponding to the coordinate value is set as the pixel position of interest, and the image data of each peripheral pixel position including the pixel position of interest is set. And an image correlation means for detecting a correlation value between the stored data stored in the image memory and a predetermined angle correction amount and a predetermined angle correction amount preset from the plurality of correlation values obtained for the predetermined angle. It is preferable to include a correction amount calculation means for generating a correction value including the position correction amount of. Further, in this case, the positional deviation detection means makes a plurality of pixels only when all the peripheral pixel positions including the target pixel position obtained for each rotation angle are overlapping scanning regions at the pixel position to be correlated. It is preferable to generate a correlation value of Further, in this case, it is preferable that the image correlation means generate a plurality of correlation values obtained for each rotation angle. Further, in this case, the image data and the scanning position are the line image sensor for reading the image,
Inputting is performed from an input device having position detecting means for detecting the reading position of the line image sensor by two detecting means, and the coordinate rotation converting means converts the scanning position coordinates of both ends of the line image sensor into plural directions on the coordinate. Rotation conversion is preferably performed in scan line units. Further, in this case, the image data and the scanning position are input from an input device equipped with a line image sensor for reading an image and a position detecting means for detecting the reading position of the line image sensor by two detecting means, The image correlation means
It is preferable to generate a plurality of correlation values obtained for each rotation angle of each scan line.

[0022]

According to the configuration of the apparatus of the present invention, the image data read by scanning the original image and the scanning position corresponding to the image data are sequentially input, and the image data is imaged based on the scanning position. An image processing device for storing in a memory, wherein a positional deviation of the scanning position is detected from the image data and the stored data stored in the image memory in an overlapping scanning area for overlapping scanning, and the positional deviation is detected. Displacement detecting means for outputting a correction value for correcting the scanning position, correction means for correcting the scanning position based on the correction value, and mapping means for storing the image data in the image memory based on the corrected scanning position. By including at least, even if the input scanning position includes a position error,
The positional deviation of the scanning position can be detected from the image information of the image data and the stored data in the overlapping scanning area in which the scanning is performed redundantly, and the positional deviation can be corrected. Then, by storing the image data in the image memory based on the corrected scanning position, the image connection deviation caused by the position error of the scanning position is prevented, and the image distortion of the planar image connected in the image memory is reduced. can do.

The configuration of the apparatus of the present invention further comprises a correction control means for changing the correction value at a predetermined ratio, and the correction means corrects the scanning position based on the correction value changed at the predetermined ratio. According to the preferable example, the correction value for correcting the positional deviation is variably controlled at a predetermined ratio, and the image data can be stored in the image memory based on the corrected scanning position. Occurrence of an area where image data is not stored in the memory is suppressed. In addition, since the scanning position is not rapidly corrected, the detection range of the correlation value when detecting the positional deviation from the image information of the image data and the stored data can be reduced, so that the circuit can be downsized. Furthermore,
Since the occurrence of a region (image missing) in the image memory where the image data is not stored is prevented, the error of the correlation value when detecting the positional deviation from the image information of the image data and the stored data in the overlapping scanning region Can be made smaller, so that the accuracy of positional deviation detection is improved.

Further, in the configuration of the apparatus of the present invention, a delay means for delaying the corrected scanning position and the corresponding image data is further provided, and the mapping means generates the corresponding image data based on the delayed scanning position. According to a preferred example in which the position where the positional deviation of the scanning position is detected and the storage position where the image data is stored in the image memory are made different by storing in the image memory, the position where the positional deviation is detected is determined by the image sensor. By making the movement direction precede, it is possible to improve the positional deviation correction accuracy in real time processing. By the way, in the real-time processing, when the image data is stored in the image memory (mapping processing) and the displacement detection is performed in parallel, pixels newly mapped and stored around the pixel of interest frequently occur. Therefore, the correlation value is affected by the new mapped pixel. Also,
If the detection position moves back and forth due to camera shake during manual scanning, the influence of the new mapped pixel becomes greater, and the detection accuracy of the positional deviation decreases. However, as described above, the storage position (mapping position) for storing the image data in the image memory is made different from the position where the positional deviation is detected, and the position where the positional deviation is detected precedes the moving direction of the image sensor. It is possible to suppress the influence of the new mapped pixel and improve the detection accuracy of the positional deviation.

Further, in the structure of the apparatus of the present invention, the positional deviation detecting means includes a scanning area holding means for sequentially storing the scanning confirmation information of the image data input to the image memory in the scanning confirmation information memory, and the scanning confirmation information. According to the preferable example of including the overlapping area detection unit that detects the overlapping scanning areas that are redundantly scanned by detecting the scanning confirmation information of the memory, it is possible to specify the overlapping scanning areas that are redundantly read. Therefore, the positional deviation can be detected only from the image data and the stored data in the overlap scanning area. Further, in this case, according to a preferable example in which the address of the image data stored in the image memory and the address of the scanning confirmation information stored in the scanning confirmation information memory are shared, the address generation mechanism and the memory control unit are simplified. Further, the access speed can be increased because access can be made simultaneously and access can be made simultaneously. Further, in this case, there is further provided memory control means for controlling the storage of the image data in the image memory according to the determination of the scanning area performed by the overlapping area detection means, and the memory control means is provided for the overlapping scanning areas to be scanned in duplicate. According to a preferable example of prohibiting the storage of the image data in the image memory when it is determined that the image data is stored, the memory control unit stores the image data of only the newly scanned area in the image memory to perform scanning. Since image data having a small position error can be preferentially stored in the image memory in an overlapping manner, distortion of the composite image is reduced. If the positional deviation is detected based on the stored image information, the detection accuracy is improved.

Further, in the configuration of the apparatus of the present invention, the positional deviation detecting means determines the storage address of the image memory corresponding to the coordinate value on the original of each pixel in the image data calculated based on the scanning position as the pixel of interest. A position and an image correlating means for detecting a correlation value between the image data and the stored data stored in the image memory with respect to the peripheral pixel positions including the target pixel position, and a plurality of the correlation values. According to a preferable example including a correction amount calculation unit that generates a preset correction value, the positional deviation detection unit can detect positional deviation in a plurality of directions by using a plurality of correlation values. . Further, in this case, the positional deviation detection means generates a plurality of correlation values only in the case where the pixel positions to be correlated are all the surrounding pixel positions including the pixel position of interest in the overlapping scanning region. According to this, since all the correlation values can be detected under the same condition, the positional deviation detection accuracy using a plurality of correlation values is improved.

Further, in the structure of the apparatus of the present invention, the correction means includes a holding means for generating a cumulative correction value by accumulating the correction values output from the positional deviation detection means, and the scanning position is corrected by the cumulative correction value. According to the preferable example described above, by accumulating the correction values for eliminating the positional deviation, it is possible to realize the correction processing of the scanning position including the accumulated error with a small circuit. This is because the misregistration detection range can be reduced because the misregistration is detected around the accumulated correction position.

Further, in the configuration of the apparatus of the present invention, the mapping means includes an address generating means for generating a storage address of the image memory corresponding to the coordinate value of each pixel in the image data on the document based on the scanning position. According to a preferred example of comprising a data storage unit for storing each pixel in the image data at the storage address of the image memory, the address generation unit determines the storage address based on the coordinates of the scanning position on the document. Since it is generated, the corresponding image data can be stored at a predetermined position in the image memory. Further, in this case, the mapping means is a pixel density conversion means for densifying the image data from adjacent pixels by interpolation, and a real number on the original of each pixel in the densified image data based on the scanning position. A unit for calculating coordinate values in units of the predetermined pixel density; an integer unit for converting real number coordinate values on the original document into integer coordinate values in units of the predetermined pixel density; A means for calculating a coordinate error value generated at the time of conversion into an integer, a means for converting the integer coordinate value into an address of an image memory, and a means for comparing the magnitude of the coordinate error value with a predetermined value. According to a preferred example in which only the pixels whose magnitude of the coordinate error value is equal to or smaller than the predetermined value are stored in the address of the image memory, the pixel data having a small error with respect to the storage address is stored. It is possible to store and select the data in the image memory, thereby improving the reproducibility quality image stored. If this stored data is used for detecting the displacement, a connected composite image with small distortion can be obtained, and the quality of the planar image is improved. In this case, the pixel density conversion means further includes means for generating an interpolated pixel by using the pixel to be processed and adjacent pixels around the pixel to be processed, and the means converts the image data to double the pixel density. For example, by increasing the pixel density, the quality of image reproduction is improved. Furthermore, since the interpolation process can be realized by a power of 2 (2 ⁿ ), the circuit can be configured by bit shift, so that the circuit can be simplified. Then, in this case, the pixel to be processed P _{i, j} (however, in the image data, the j-th pixel data of the image data on the i-th line is shown) and its neighboring pixels P _{i, j + 1} , P _{i. -1, j} , P _{i-1, j + 1} , the interpolation pixels Q _{i, j} , R _{i, j} , S _{i, j} are calculated from the above (Equation 1).
According to the preferable example of calculating, the predetermined interpolation pixel can be easily generated. Further, in this case, the predetermined value is 0.3 to 0.3 with the predetermined pixel density as a unit.
According to the preferable example of 0.35, it is possible to eliminate mapping missing pixels.

Further, in the configuration of the device of the present invention, the image data and the scanning position include an image sensor for reading an image, and an input device provided with position detecting means for detecting the reading position of the image sensor by two detecting means. According to the preferable example of inputting from, the scanning position of the image sensor can be specified by the two detecting means, so that a large original can be read by reciprocating one-stroke writing. Further, in this case, according to a preferable example in which the image sensor is a line sensor, since the document can be scanned line by line, the size can be easily reduced. Further, in this case, according to a preferable example in which the input device is a hand scanner,
Since the desired position and area on the original can be freely input, operability is improved. Further, in this case, according to a preferable example in which the detection unit is configured by a wheel that rotates in contact with the document surface and an encoder that outputs a pulse corresponding to the rotation of the wheel, the tablet or the reference grid is Compared to a hand scanner that uses a printed sheet as an auxiliary device, an expensive auxiliary device is not required, so that it is possible to realize an image processing device that is small in size and excellent in portability and operability. In this case, further, according to the preferable example in which the encoder outputs pulses of two systems having different phases, the encoder can detect the rotation direction and the moving distance of the wheel by counting the pulses of two systems having different phases. Therefore, it is possible to detect the position of the wheel after the movement based on the predetermined position. Also in this case,
According to a preferable example in which the input device is detachably attached to the main body of the portable information device, a small-sized input device excellent in operability is realized.

Further, in the structure of the device of the present invention, there is further provided coordinate rotation conversion means for rotating the scanning position at a plurality of predetermined angles about predetermined coordinates, and the position deviation detection means is provided at a plurality of predetermined angles. The coordinate value of each pixel in the image data on the original is calculated based on the coordinates of the scanning position corresponding to the pixel position, and the storage address of the image memory corresponding to the coordinate value is set as the pixel position of interest. Image correlating means for detecting a correlation value between the image data and the stored data stored in the image memory for each of its peripheral pixel positions, and a plurality of the correlations obtained for the predetermined angle, respectively. According to a preferable example in which a predetermined angle correction amount preset from the value and a correction amount calculation means for generating a correction value including a predetermined position correction amount are provided, It is possible to detect the positional deviation based on the rotation converted scan position coordinate by the conversion means, it is possible to detect the positional deviation in the rotational direction of the image sensor. As a result, the image quality of the flat image connected in the image memory is improved. Further, in this case, the positional deviation detecting means, at the pixel position to be correlated, only if the surrounding pixel positions including the target pixel position obtained for each rotation angle are overlapping scanning regions, a plurality of correlations are detected. According to the preferable example of generating the value, all the correlation values can be detected under the same condition, so that the positional deviation detection accuracy using a plurality of correlation values is improved. Further, in this case, according to a preferable example in which the image correlation unit generates a plurality of correlation values obtained for each rotation angle, the rotation direction shift is generated by generating the correlation value for each rotation angle. It is possible to obtain a correlation value for detecting Further, in this case, the image data and the scanning position are input from an input device equipped with a line image sensor for reading an image and a position detecting means for detecting the reading position of the line image sensor by means of a coordinate rotation. According to a preferred example in which the conversion unit rotationally converts the scanning position coordinates of both ends of the line image sensor in a plurality of directions on the coordinate, by rotationally converting both end positions of the line image sensor on the coordinate, Since the positional deviation in the rotational direction can be detected for each scanning line, the positional deviation in the rotational direction can be detected by a simple circuit. Further, in this case, the image data and the scanning position are inputted from an input device equipped with a line image sensor for reading an image and a position detecting means for detecting the reading position of the line image sensor, and the image correlation is obtained. According to a preferred example in which the means generates a plurality of correlation values obtained for each rotation angle of each scanning line, by detecting the positional deviation in the rotation direction for each scanning line, the line image on the coordinate is detected. Both ends of the sensor can be accurately corrected with a simple circuit. As a result, it is possible to use a line image sensor capable of miniaturization and high resolution as an input device.

[0031]

EXAMPLES The present invention will be described in more detail below with reference to examples. <First Embodiment> First, an image processing apparatus according to a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an image processing apparatus and an image reading unit for reading and scanning an original image according to a first embodiment of the present invention. In FIG. 1, A indicates an image processing apparatus and B indicates an image reading unit.

As shown in FIG. 1, the line image sensor 1 generates image data by manually scanning the original 9 and reading the original image. Then, the generated image data is output to the image buffer 4.

As shown in FIG. 16, the image reading section B for reading and scanning the original image is the line image sensor 1
And the case 104 of the hand scanner body to which is attached,
Two wheels 31, 32 provided at both ends of the line image sensor 1 and an encoder 2 attached to each wheel 31, 32 for detecting the rotation of the wheels 31, 32
a and 2b. Each encoder 2a, 2b
Two-phase A-phase pulse and B-phase pulse having different phases by 90 degrees are generated according to the rotation angles of the wheels 31 and 32. Then, using the A-phase pulse and the B-phase pulse, the wheels 31,
The rotation direction of 32 is detected. Since the A-phase pulse and the B-phase pulse have a phase difference of 90 degrees, the level of the A-phase pulse detected by the rising edge of the B-phase pulse is
The “H” level and the “L” level are identified according to the rotation directions of 1, 32. The "L" level of the identified signals 341 and 361 is set to the forward direction of the wheels 31 and 32 (UP direction).
Then, the "H" level is in the reverse direction (DOWN direction) of the wheels 31, 32. The position counters 33 and 35 are
When the signals 341 and 361 are at “L” level, the count value is increased according to the number of B-phase pulses, and the signals 341 and 36 are increased.
When 1 is "H" level, the count value is decreased according to the number of B-phase pulses. The position coordinate detection circuit 37 inputs the count values 331 and 351 from the position counters 33 and 35, and considers the rotation directions of the wheels 31 and 32.
The moving distances of 1 and 32 are detected.

Next, as shown in FIG. 1, the scanning position detecting circuit 3 determines each wheel 3 based on the moving distance of the wheels 31, 32.
The coordinates of 1 and 32 on the original 9 are calculated. further,
The scanning position detection circuit 3 converts the coordinates of the wheels 31 and 32 into the coordinates of the read pixels at both ends of the line image sensor 1 and outputs the coordinates as scanning position coordinates 300 to the misregistration detection circuit 7. As shown in FIG. 16, the scanning position detection circuit 3
Is a position coordinate detection circuit 37, position counters 33, 35,
However, details of the operation will be described later.

The mapping circuit 5 performs pixel density conversion on the image data 400 output from the image buffer 4 and outputs densified image data 500. The positional shift detection circuit 7 calculates a correlation value between the high-density image data 500 output from the mapping circuit 5 and the stored data stored in the image memory 6. Further, the position shift detection circuit 7 uses the position correction amount calculated based on this correlation value to scan position coordinate 300.
Is corrected and output to the mapping circuit 5 as corrected position coordinates 710. Further, the mapping circuit 5 uses the corrected position coordinates 710 output from the position shift detection circuit 7 to generate the image memory 6
Generates the address of. Then, the mapping circuit 5 stores each pixel data of the densified image data 500 in the image memory 6 via the bus 600. Details of the operations of the mapping circuit 5 and the positional deviation detection circuit 7 will be described later.

Next, the operation of the scanning position detecting circuit 3 will be described in more detail. FIG. 2 is an operation explanatory diagram of the scanning position detection circuit. In FIG. 2, the thick line indicates two wheels 3
The movement loci of 1 and 32 are shown. The coordinates indicating the positions of the two wheels 31, 32 when the line image sensor 1 (FIG. 1) reads the pixel data of the i-th line are P0 _i (X0 _i , Y0 _i ) and P1 _i (X1 _i , respectively). , Y
1 _i ). Now, assuming that the coordinates of P0 _i-1 and P1 _i-1 are known, the coordinates of P0 _i and P1 _i are approximately calculated using the following (Equation 2).

[0037]

[Equation 2]

Here, · is an operation for multiplication and / is an operation for division. Hereafter, * is an operator indicating multiplication,
/ Is used as an operator indicating division. L0 _i-1 is the distance traveled by the wheels 31 and 32 from the start of reading to the time of reading the (i-1) th line. △ L0
_i is the distance traveled by the wheels 31 and 32 from the start of reading the (i-1) th line to the time of reading the i-th line. The moving distance can be a negative value because the rotation direction of the wheels 31 and 32 is taken into account. Further, the moving distance of the wheels 31 and 32 on the original 9 should be calculated as P × N by using the pulse number N of the encoders 2a and 2b and the resolution P (inch / 1 pulse) per pulse shown in FIG. Obtained by Position coordinate detection circuit 37
Indicates the count values 331, 35 of the position counters 33, 35.
1 in synchronism with the reading cycle of the line image sensor 1, and based on the difference between the count values detected on the i-th line and the (i-1) -th line, the moving distance on the document 9 including the rotation directions of the wheels 31 and 32. ΔL0 _i is detected. D
Is the distance between the wheel 31 and the wheel 32. In the above (Equation 1), Δθ = | θ _i −θ _i−1 | = | ΔL0 _i −ΔL1
_This is an approximate calculation with _i | / D set to 0. Δθ is a change angle of the line image sensor 1 during one line scanning time of the line image sensor 1. By using the above (Equation 1), the two wheels 3 at the start of reading
If the coordinates of 1 and 32 are decided, the two wheels 31 and 32
The coordinates can be calculated from the moving distance of the.

FIG. 3 is an explanatory diagram in the case of calculating the coordinates of the read pixels at both ends of the line image sensor. The coordinates of the wheel 31 are P0 (X0, Y0), and the coordinates of the wheel 32 are P
1 (X1, Y1). The coordinates Ps (Xs, Ys) and Pe (Xe,
Ye) is calculated by the following (Equation 3).

[0040]

(Equation 3)

Here, D is the distance between the wheel 31 and the wheel 32, d1 is the distance from the wheel 31 to the read pixel Ps, and d2 is the distance from the wheel 31 to the read pixel Pe. The scanning position detection circuit 3 uses the moving distances of the wheels 31 and 32 obtained from the two-phase pulses generated by the encoders 2a and 2b to perform the calculations of (Formula 2) and (Formula 3), and the line image sensor 1 The coordinates Ps (Xs, Ys) and Pe (Xe, Ye) of the read pixels at both ends of are output to the misregistration detection circuit 7 as scanning position coordinates 300.

FIG. 4 is an explanatory diagram of the scanning area of the line image sensor. Referring to FIG. 4, when the reading area width of the original 9 is larger than the length of the line image sensor 1,
The movement of the line image sensor 1 by manual scanning will be described. In order to read the original 9, the operator brings the hand scanner body into contact with the original 9 and performs manual scanning while reciprocating on the original 9. At this time, the two wheels 31, 32 attached to the hand scanner body rotate,
Two-phase pulses are output from the encoders 2a and 2b. FIG. 4 shows a reading area on the original 9 read by the line image sensor 1.

Since the line image sensor 1 cannot scan the entire width of the original 9, the image reading section B (FIG. 1) reads the entire original 9 by the reciprocating motion of the line image sensor 1. Although the positions of only the pixels at both ends of the line image sensor 1 are shown in FIG. 4, the line image sensor 1 reads an image on a line connecting the pixels at both ends. For example, when the pixels at both ends of the line image sensor 1 are points A and B, respectively, the line image sensor 1 reads on the line connecting the points A and B (hereinafter, this is referred to as "reading position AB"). It is written.).

In FIG. 4, the line image sensor 1
Scans to the reading position CD with the reading position AB as the scanning start position. Each pixel in the image data obtained by reading the area ABCD surrounded by the points A, B, D, and C is based on the scanning position coordinate 300 output from the scanning position detection circuit 3 (FIG. 1), and the mapping circuit 5 It is newly stored in the image memory 6 by (FIG. 1). Hereinafter, such a region will be referred to as a "new scan region".

Next, the line image sensor 1 moves in the returning direction and moves from the reading position CD to the reading position E-.
Scan to F. An area CDGE surrounded by points C, D, G, and E is an area where an image is read in an overlapping manner. Hereinafter, the area that is read redundantly is referred to as an “overlapping scanning area”. An area DGF surrounded by points D, G, and F is a new scanning area. in this way,
There are three scanning areas: an overlapping scanning area CDGE, a new scanning area ABGEC, and a new scanning area DFG.

If there is no position error in the scanning position coordinate 300, each pixel in the read image data can be mapped and stored in the image memory 6 based on the scanning position coordinate 300. That is, even if the read image data of the overlapping scanning area CDGE is overwritten in the image memory 6, the new scanning area ABGEC and the overlapping scanning area CDGE
At the seam portion, the read image in the image memory 6 is not displaced. However, the mechanism design accuracy of the hand scanner, the slip between the wheels 31 and 32 and the original document 9, the sinking of the wheels 31 and 32 into the original document 9 and the width between the wheels 31 and 32 during the manual scanning of the curve. Due to the influence and the like, the scanning position coordinate 300 includes a position error. Further, the scanning position detection circuit 3 includes encoders 2a, 2
Since the two-phase pulse output from b is counted to obtain the moving distance of the encoders 2a and 2b, the position error is accumulated. Therefore, when the image data 400 is mapped to the image memory 6 using the scanning position coordinate 300, the image shift occurs at the joint portion. Here, “mapping” refers to the operation of storing the read image data at a predetermined address of the image memory 6.

In order to eliminate this image shift, the position shift detection circuit 7 uses the image data stored in the image memory 6 of the overlap scanning area CDGE and the high density image data 500 to show the degree of correlation between them. Calculate the correlation value.
Further, the position shift detection circuit 7 calculates a position correction amount for correcting the scanning position coordinates 300 based on the correlation value. The position shift detection circuit 7 corrects the scanning position coordinate 300 according to the position correction amount, and the corrected position coordinate 710
Is output to the mapping circuit 5. The mapping circuit 5 generates an address for mapping each pixel in the densified image data 500 in the image memory 6 according to the corrected position coordinate 710, and stores the address in the image memory 6. The extraction of the overlap scanning area CDGE will be described later.

FIG. 7 is an explanatory diagram of the image memory. The bit of each pixel of the image memory 6 is composed of a storage bit (bit 7) of a write flag that holds scanning confirmation information and a storage bit (bits 0 to 6) of image data. Here, the number of storage bits of the image data is not specified and can be designed according to the required number of gradations. In this embodiment, an image with 128 gradations is handled, and 0
To store grayscale data having values from 1 to 127,
7 bits per pixel are secured in the image memory 6. The write flag of bit 7 indicates that the image data is the image memory 6
If the image data has not been written to (i.e., in the unstored state), it is "0", and if the image data has already been written (i.e., in the stored state), it is "1". By thus sharing the memory for storing the image data and the memory for storing the scanning confirmation information, the address generation and the memory control unit can be simplified, and further,
Since it is possible to access at the same time, the access speed can be increased. It should be noted that the memory for storing the image data and the memory for storing the scanning confirmation information may be separately configured, and the scanning confirmation information may be stored in block units instead of pixel units. It is only necessary to be able to store the scanning confirmation information that can specify the storage position of the image data.

Next, the operation of the displacement detection circuit 7 will be described. FIG. 5 is a block diagram showing a position shift detection circuit. Before the reading scanning of the line image sensor 1 is started, all the data in the image memory 6, the position correction amount 703 of the correction amount calculation circuit 73, and the image correlation circuit 72.
The correlation table in is initialized to "0". After this initialization, the scanning position coordinates 300 are corrected by the position correction circuit 74 every time a line is scanned by the line image sensor 1 and output to the mapping circuit 5 as corrected position coordinates 710. At the time when the reading scan of the line image sensor 1 is started, the position correction amount 703 is “0”.
Therefore, the scanning position coordinate 300 and the correction position coordinate 710
Have the same coordinate values.

The mapping circuit 5 densifies the image data 400 by a pixel density conversion process, and densifies the image data 400.
00 is generated. Further, the mapping circuit 5 uses the input corrected position coordinates 710 to densify the image data 500.
Storage address A of each pixel data Pn of
Calculate DRn. The mapping circuit 5 reads the pixel data Pn stored in the address ADRn of the image memory 6 via the bus 600 according to the address ADRn,
This pixel data Pn is output to the overlap area detection circuit 71 via the bus 600. The details of the operation of the mapping circuit 5 will be described later.

The overlap area detecting circuit 71 is arranged to detect the pixel data P.
The write flag (bit 7) of n is checked to determine whether the image data has been stored at the address ADRn of this pixel data Pn. When the write flag (bit 7) of the pixel data Pn is 1, it indicates that the image data has already been stored in the address ADRn by the reading scan of the line image sensor 1, so that the pixel data Pn overlaps the scan region. Is determined to be included in. When the write flag (bit 7) of the pixel data Pn is 0, it is determined that the pixel data Pn is included in the new scanning area. Overlapping area detection circuit 71
Outputs the determination signal 701 to the image correlation circuit 72 and the mapping circuit 5. Here, the determination signal 701 is the pixel data Pn
Is "0" when included in the new scanning area, and "1" when included in the overlapping scanning area.

The image correlation circuit 72 performs a correlation value calculation process for the pixel data Pn when the determination signal 701 is "1", and a correlation value calculation for the pixel data Pn when the determination signal 701 is "0". Do not process. The mapping circuit 5 stores the densified pixel data Pn in the image memory 6 when the determination signal 701 is “0”, and stores the densified pixel data Pn when the determination signal 701 is “1”. Do not store in. This series of processing operations for each pixel is
This is performed for all pixel data of one line of the densified image data 500.

At the time when the above processing of the high density image data 500 for one line is completed, the image correlation circuit 72 creates a correlation table created by performing the correlation value calculation processing only for the pixels included in the overlap scanning area. The position shift direction of the scanning position coordinate 300 is detected by using this. further,
The image correlation circuit 72 outputs an offset value 702 for canceling the positional deviation to the correction amount calculation circuit 73.
When the densified pixels of all one line are included in the new scanning area, the correlation table of the image correlation circuit 72 remains the initial value "0". In this case, the offset value 702 is “0” (no positional deviation).

The correction amount calculation circuit 73 uses the offset value 70
2 is added to the accumulated value of the correction amount held therein, and the resultant value is output to the position correction circuit 74 as the position correction amount 703. The position correction circuit 74 adds the scanning position coordinates 300 of the image data of one line to be processed next and the position correction amount 703,
The correction position coordinates 710 are output to the mapping circuit 5. Thereafter, the above-described series of processing is sequentially repeated for each line.

Next, regarding the operation of the mapping circuit 5, FIG.
This will be described with reference to FIGS. 6 is a block diagram showing a mapping circuit, FIG. 7 is an explanatory diagram of an image memory, and FIG. 9 is an explanatory diagram of pixel density conversion.

As shown in FIG. 6, the pixel density conversion circuit 51.
Generates three interpolated pixels for each pixel data in the image data 400, and outputs the densified image data 500 that has been doubled in density.

First, FIG. 9 shows a method of generating an interpolation pixel.
Will be explained. In FIG. 9, P _{i, j} is the image data 40
The j-th pixel data of the image data of the i-th line at 0 is shown. Black dots are coordinate points of each pixel data. FIG. 9A shows four adjacent pixels in the image data 400. In FIG. 9B, Q _{i, j} , R _{i, j} and S _{i, j} are interpolation pixel data for the pixel data P _{i, j} in the image data 400. Each interpolated pixel data Q _{i, j} ,
R _{i, j} and S _{i, j} are calculated by the above (Equation 1).

Next, the coordinate value calculation circuit 52 will be described. As shown in FIG. 6, the coordinate value calculation circuit 52 receives the corrected position coordinates 710 which are the coordinate values after the correction of the pixels at both ends of the line image sensor 1. Coordinate value calculation circuit 52
Calculates the coordinate value 520 of each pixel of the densified image data 500 using the input corrected position coordinates 710. As shown in FIG. 7, both end pixels Ps of the line image sensor 1
The operation of the coordinate value calculation circuit 52 when the coordinates of _i and Pe _i (correction position coordinates 710) are (Xs _i , Ys _i ) and (Xe _i , Ye _i ) respectively will be described. The suffix i indicates that it is the corrected position coordinate of the i-th line of the image data 400. Here, the read pixel density of the line image sensor 1 is 8 pixels / mm, and the image memory 6
The pixel density of the image stored in is set to 8 pixels / mm. Xs
_i , Ys _i , Xe _i, and Ye _i are real values in units of 1/8 mm.

When the number of read pixels in one line of the line image sensor 1 is Nd and the pixel number in one line is j _, the coordinates (XP _{i, j} , YP _{i, j} ) of the pixel data P _{i, j}
Is calculated using the following (Equation 4).

[0060]

[Equation 4]

Pixel data P_{i, j}Interpolated images corresponding to
Raw data Q_{i, j}, R_{i, j}And S_{i, j}Coordinates (XQ_{i, j},
YQ_{i, j}), (XR_{i, j}, YR_{i, j}) And (XS_{i, j},
YS _{i, j}) Is calculated using the following (Equation 5).

[0062]

(Equation 5)

The coordinate value calculation circuit 52 calculates the coordinate value 520 of each pixel in the densified image data 500 by performing the arithmetic processing shown in (Equation 4) and (Equation 5).
The integer conversion circuit 53 converts the real-valued coordinate value 520 into an integer and outputs an integer-converted coordinate value 530. Real number coordinate value 52
0 for (X _real , Y _real ), and the integer coordinate value 530 for (X _real
int , Y _int ), the integer coordinate value is calculated using the following (Equation 6).

[0064]

(Equation 6)

In [Equation 6], [] indicates an operation of rounding off the decimal point. Performing the fractional part truncation process after adding 0.5 is equivalent to rounding off.

The address generation circuit 54 is an integer circuit 53.
The converted integer coordinate value 530 is converted into the address 540 of the image memory 6. FIG. 10 shows the address arrangement of the image memory. The image memory 6 has M pixels in the X direction and Y pixels.
The page memory has N pixels in the direction. The address of the upper left pixel of the image memory 6 is 0, and the address of the upper right pixel is (M
-1), the address of the lower right pixel is (MN-1).
If the integer coordinate value 530 is (X _int , Y _int ), the address ADR of the image memory 6 is calculated by the following (Equation 7).

[0067]

(Equation 7)

The error calculation circuit 55 has a real coordinate value 52.
0 and the integer coordinate value 530 are input, and the real coordinate value 52
The coordinate error 550 caused by converting 0 into an integer is output to the comparison circuit 56. If the coordinate error in the X direction is Ex and the coordinate error in the Y direction is Ey, the coordinate error (Ex, Ey) is
It is calculated by the following (Equation 8).

[0069]

(Equation 8)

Here, || indicates an operation that takes an absolute value. Hereinafter, || is used as an operator that takes an absolute value. Ex and Ey take values of 0 to 0.5. The comparison circuit 56 compares the coordinate errors Ex and Ey with a predetermined value. The comparison circuit 56 calculates the coordinate errors Ex and Ey
When both are smaller than the predetermined value, a signal 560 of “1” is output to the access circuit 57.

The access circuit 57 accesses the image memory 6 (FIG. 5) via the bus 600. The address of the image memory 6 is transferred from the address generation circuit 54 to the access circuit 5
7 is designated by the address 540 output to No. 7. The access circuit 57 stores the high-density image data 500 in the image memory 6 only when the determination signal 701 is “0” and the signal 560 is “1”. That is, the mapping of an arbitrary pixel in the high-density image data 500 onto the image memory 6 is a condition that the pixel is a pixel included in the new scanning area and the coordinate errors Ex and Ey are both smaller than the predetermined value. It is performed only when the condition is met. Pixels that do not satisfy this condition are not mapped to the image memory 6. By performing memory control so that new image data is not stored in the stored area on the image memory 6, even if a scanning position including a cumulative position error is input, when the scanning is repeated by manual reciprocating one-stroke writing In addition, since the image data having a small cumulative position error is preferentially stored in the image memory 6, the image memory 6 always stores an image having a small scanning position error. If this stored data is used for detecting the displacement, a connected composite image with small distortion can be obtained, and the quality of the planar image is improved.

FIG. 11 is an explanatory view of the mapping operation of the high density image data onto the image memory. FIG. 11A shows high-density image data 500. In FIG. 11A, the black dots are the coordinate values of the pixels P, Q, R, and S. The pixel density of the high-density image data 500 is a minimum of 16 pixels / m
m. FIG. 11B shows pixels of the image memory 6. In FIG. 11B, the black dot is the coordinate value of the pixel W. The distance U indicates a predetermined value used in the comparison circuit 56 of the mapping circuit 5. The image memory 6 stores image data having a pixel density of 8 pixels / mm. Figure 11
(C) is the high-density image data 500 (FIG. 11A).
And pixels of the image memory 6 (FIG. 11B) are overlapped in the same coordinate system. In the case of FIG. 11C, since the coordinate values of each pixel P, Q, R, and S of the high-density image data 500 are outside the area T, all the pixels P, Q, R, and S are in the image memory. It is not mapped to pixel W of 6. That is, in the image memory 6, there are pixels that are not imaged even though they are the original reading area (pixels that are missing images). The mapping missing pixels can be eliminated by expanding the region T. However, when the area T is widened, the coordinate error at the time of mapping becomes large, so that the distortion of the image mapped in the image memory 6 becomes large. From the viewpoint of image distortion, the smaller the area T, the better.

The upper limit value U _max of the distance U for eliminating mapping missing pixels is expressed by the following (Equation 9) with the pixel pitch of the image memory 6 as a unit.

[0074]

[Equation 9]

In the case of this embodiment, since the pixel density of the image memory 6 is 8 pixels / mm, the unit is 1/8 mm.
By setting the distance U to 0.35, it is possible to eliminate mapping missing pixels. In the case where the occurrence of a certain number of missing pixels is allowed and emphasis is placed on reducing image distortion, the distance U may be set in the range of 0.3 to 0.35. When the distance U is set to 0.3 or less, image-missing pixels occur frequently, and the image quality of a reproduced image significantly deteriorates.

Description of the operation of the positional deviation detection circuit 7 shown in FIG.
Return to FIG. 8 is an explanatory diagram of the correlation table. Image correlation
The circuit 72 will be described mainly with reference to FIG. Figure 8
8A is an explanatory diagram of a correlation position that is a target of correlation processing, and FIG.
(B) is an explanatory view of a correlation table. Position correction circuit 7
The scanning position coordinate 300 of the i-th line input in 4 is set to P1.
0 (X1, Y1), P20 (X2, Y2)
Positive amount 703 is △ Xoffset _i, △ Yoffset_iAnd position
The correction circuit 74 uses the scanning position coordinate 300 and the position correction amount 7
03, using the following (Equation 10), the corrected position coordinate 71
Calculate P11 (X3, Y3) and P21 (X4, Y4) of 0
Put out.

[0077]

[Equation 10]

The image correlation circuit 72 operates when the determination signal 701 from the overlap area detection circuit 71 is "1" (that is,
Only when the processed pixel is included in the overlapping scanning area), the correlation value is calculated for the processed pixel, and the correlation table is updated. The pixel Pn in the image memory 6 corresponding to the coordinates of the pixel to be processed is the pixel of interest. The correlation value is calculated by calculating the difference value between the pixel data in the image memory 6 and the pixel data of the pixel to be processed, which corresponds to the coordinate obtained by increasing or decreasing the coordinate of the pixel to be processed by a minute value.

The coordinates of the pixel of interest Pn are (Xn, Yn)
If the small coordinate values are Δhx and Δhy, the phase of the pixel to be processed
Coordinates (Xh_mn,
Yh _mn) Is calculated by the following (Equation 11).

[0080]

[Equation 11]

Here, m and n are -1, 0 and 1 respectively.
Takes the value of. In addition, [] indicates an operation of rounding off after the decimal point. In FIG. 8A, P12 → P22
Indicates the position of one line for calculating the correlation value when m = 1 and n = 1. Assuming that the value of the correlation table corresponding to the target coordinates for calculating the correlation value is h (m, n), FIG.
A correlation table shown in (b) is created.

If the pixel number in one line of the high-density image data 500 is j, the data value is Dn _j , and the pixel data for which the correlation value is to be calculated in the image memory 6 is Dh _jmn ,
The value h (m, n) of each correlation table is calculated by the following (Equation 12).

[0083]

(Equation 12)

Here, ho (m, n) is the pixel number (j
It is a value of the correlation value table generated by the correlation value calculation up to -1). Before starting the correlation value calculation for one line, all the values in the correlation table are initialized to "0".

The image correlation circuit 72 completes the correlation table by performing the above correlation value calculation for all the pixels of one line in the densified image data 500. The target coordinates for calculating the correlation value are the corrected position coordinates 710 calculated by the above (Equation 10).

The image correlation circuit 72 retrieves (m _min , n _min ) holding the minimum value of h (m, n) at the time when the calculation of the correlation value of one line is completed, and corrects it as the offset value 702. It is output to the amount calculation circuit 73. There are a plurality of minimum values in the correlation table, and the minimum value is (m _min ,
n _min ) = (0,0) is included, (0,0)
The minimum value of is used preferentially. The fact that the correlation value h (m _min , _nmin ) in the correlation table is the smallest means that
When a minute value of (Δhx × m _min , Δhy × n _min ) is added to the coordinates of each pixel for mapping, the image in the image memory 6 and the image of the line to be mapped best match. It shows that you do. Also, when there are multiple minimum values and the minimum value includes the center of the correlation window,
The offset value 702 is “0”. For example, if a 3 × 3 correlation window is set, h (0,0) becomes the center of the correlation window. Here, m and n take values of −1, 0, and 1, respectively.

The correction amount calculation circuit 73 uses the offset value 70
The calculation shown in the following (Equation 13) is performed using 2 (m _min , _nmin ). Here, the offset value 702 is defined as △ X = △ h
Let x × m _min and ΔY = Δhy × n _min .

[0088]

(Equation 13)

In the above (Equation 13), the suffix i
Represents the position correction amount 703 when the correlation table of the i-th line of the high-density image data 500 is completed. The position correction circuit 74 displays (ΔXoffset _i ,
Scan position coordinate 3 by adding ΔY offset _i ).
00 is corrected and output as corrected position coordinates 710.

As described above, according to the configuration of the image processing apparatus of the present embodiment, even if the input scanning position includes a positional error, the image data and the stored data in the overlapping scanning area that are scanned in duplicate are stored. The positional deviation of the scanning position can be detected from the image information of and to correct the positional deviation. And
By storing the image data in the image memory 6 based on the corrected scanning position, connection deviation of images caused by a positional error of the scanning position is prevented, and image distortion of the planar image connected in the image memory 6 is reduced. can do.

In the present embodiment, the correlation target pixel positions are nine, but the positions are not necessarily limited to nine. When the amount of displacement is large, the number of correlation target pixel positions may be increased to widen the correlation range. Alternatively, the correlation table may be created at N (N> 1) scanning line intervals and the positional deviation may be corrected at N (N> 1) scanning line intervals. For example, N = 8 may be set.

<Second Embodiment> Next, as a second embodiment of the present invention, a description will be given of the processing when the offset value 702 corresponding to the positional deviation detected by the image correlation circuit 72 (FIG. 5) is large. To do.

When the offset value 702 corresponding to the positional deviation detected by the image correlation circuit 72 is large, the position correction amount 703 greatly changes at the position where the offset value 702 is added, and as a result, the corrected position coordinate 710 is corrected. The position changes greatly.

When the corrected position coordinate 710 changes significantly
15 shows the data storage state in the image memory 6 of FIG.
It demonstrates using. FIG. 15 shows the data in the image memory.
It is explanatory drawing which shows a storage state. In FIG. 15, (i-1)
The i-th line is used as the line for image correlation detection
Is the line for position correction. In addition, (i-1)
Ps for both end pixels_i-1, Pe_i-1, I-th line both
Edge pixel is Ps_i, Pe _iIt has been described as.

As shown in FIG. 15, according to the correction position coordinate 710 at the position where the amount to be corrected greatly changes,
When the mapping circuit 5 stores image data in the image memory 6,
An area in which no data is stored (a mapping missing area due to position correction) occurs between the (i-1) th line and the i-th line.

A method of suppressing the generation of the area in which the data is not stored will be described with reference to FIGS. 12, 13 and 14. FIG. 12 is a block diagram of a positional deviation detection circuit including a division control circuit, FIG. 13 is a timing chart of the division control circuit, and FIG. 14 is a diagram for explaining the difference between the position correction amount and the division position correction amount.

12 is different from the block diagram of the positional deviation detection circuit 7 in FIG. 5 in that the offset value 702 of the image correlation circuit 72 is not directly input to the correction amount calculation circuit 73. The offset value 702 is a division offset value 70 that suppresses the occurrence of an area (image missing area due to position correction) in which data is not stored by the division control circuit 75.
After being converted into 4, it is input to the correction amount calculation circuit 73.

A parameter NB for dividing the offset value 702 is held in the register 754 in order to suppress the occurrence of the mapping missing area due to the position correction. The offset value 702 is divided by the divider 751 by the division parameter NB held in the register 754, and the switch 7
52 is a divider 7 depending on the control output of the switch control 753.
Turn ON / OFF the output of 51. When “ON”, the division offset value 704 is output from the divider 751.
In the case of “OFF”, the division offset value 704 is not output (no offset value). If the division parameter NB is 2 ⁿ , the divider 751 can be configured by n-bit shift. The division parameter NB is not specified and can be arbitrarily set according to the maximum value that the offset value 702 can take.

The operation of the division control circuit 75 will be described with reference to FIG. 13 according to a timing chart. Here, the division parameter NB will be described as 4. Further, it is assumed that the line position where the image correlation is first performed is i-th and the image correlation detection is performed at intervals of 8 scanning lines. In FIG. 13, only the number is described such that the i-th line is 0 (line position) and the (i + 8) -th line is 8 (line position).

The mapping circuit 5 outputs the line synchronization signal of one line to the division control circuit 75 and the image correlation circuit 72 via the bus 600 when all the image signals of one line read by the line sensor 1 are mapped. To do.

The image correlation circuit 72 sets the first line position where the scanning is started to the initialization line position (i = 0 line position), and counts the line synchronization signal from the initialization line position. Then, the remainder obtained by dividing the count number i by 8 is 0.
In this case, the image correlation detection is performed (excluding the initialization line position of i = 0), and the offset value 702 is updated in synchronization with the line synchronization signal before the reading of the next line position is started. When the image correlation detection ends, the image correlation circuit 72 outputs the correlation end signal to the switch control 753 of the division control circuit 75 via the bus 600. The switch control 753 sets the count value of the next line position at which the correlation end signal of the image correlation circuit 72 is received, and starts counting the line synchronization signal from the next line position. In response to the division parameter NB held in the register 754, the switch control 753 turns ON the switch 752 when “1 ≦ count value ≦ 4” and “count value> 4”.
In the case of, the switch 752 is turned off.

The correction amount calculation circuit 73 inputs the divided division offset value 704, and receives the position correction amount 70 for one line before.
The accumulated value of the register 731 storing 5 and the division offset value 704 are added by the adder 732, and output to the position correction circuit 74 as a position correction amount 705 (hereinafter referred to as “division position correction amount”) controlled by division.

Next, in the correction amount calculation circuit 73, the position correction amount 703 generated by inputting the offset value 702 is input.
The difference between (without division control, FIG. 5) and division position correction amount 705 (with division control) generated by inputting the division offset value 704 will be described with reference to FIG.

Similar to FIG. 13, it is assumed that the line position where the image correlation is first performed is i-th and the image correlation is detected at the intervals of 8 scanning lines. In FIG. 14, only the numbers are described such that the i-th line is 0 (line position) and the (i + 8) -th line is 8 (line position). Further, since the offset value 702 for canceling the positional deviation is the same processing in both the X direction and the Y direction, only one of the X directions is described. Also, at the i-th line position,
It is assumed that the cumulative value of the register 731 of the correction amount calculation circuit 73 is “0” and the number of divisions of the register 754 of the division control circuit 75 is 4. Incidentally, when the correlation window of 3 × 3 is set, the number of divisions is 4 as described above, and when the correlation window of 5 × 5 is set, the number of divisions of 8 is sufficient. The “variable control amount” described in claim 3 of the present invention is the division offset value 704.

In FIG. 14, the i-th line and (i + 8)
Since the image correlation detection is performed on the line, the offset value 702 from the image correlation circuit 72 is the (i + 1) th line (offset value = Δhx × m) and the (i + 9) th line (offset value = −Δhx ×). It is output in m).

When the division control is not performed (FIG. 5), the correction amount calculation circuit 73 inputs the offset value 702, and in the next (i + 1) line, the accumulated value (cumulative value = 0) of the register 731 and the offset The position correction amount 703 is obtained by adding the value Δhx × m. Similarly, the correction amount calculation circuit 73 causes the register 731 to operate on the (i + 9) th line.
Cumulative value (cumulative value = Δhx × m) and offset value −Δ
A position correction amount 703 obtained by adding hx × m is obtained.

When performing division control, the correction amount calculation circuit 73
Inputs the division offset value 704, and the next (i + 1)
In the line, the cumulative value of the register 731 (cumulative value =
0) and the offset value Δhx × (m / 4) are added to obtain a division position correction amount 705. In addition, the correction amount calculation circuit 73
At the (i + 2) th line, the division position correction amount 705 is obtained by adding the cumulative value (cumulative value = Δhx × (m / 4)) of the register 731 and the offset value Δhx × (m / 4). After that, up to the (i + 4) th line, Δhx ×
A division position correction amount 705 that increases by (m / 4) is obtained.
From the (i + 4) th line to the (i + 8) th line, since the division offset value 704 is 0, there is no change. In addition, the correction amount calculation circuit 73 adds the accumulated value (cumulative value = Δhx × m) of the register 731 and the offset value −Δhx × (m / 4) at the (i + 9) th line, the division position correction amount. To obtain 705. After that, from the (i + 9) th line to the (i + 12) th line, −Δhx × (m / 4)
A division position correction amount 705 that is reduced by just (I + 1
3) Since the division offset value 704 becomes 0 from the line 3, it does not change.

If the large offset value 702 is detected without the division control, the position correction amount 703 is greatly increased or decreased between the (i-1) th line and the next i-th line where the image correlation detection is performed as it is. To do. For this reason, as shown in FIG. 15, an area in which no data is stored (a mapping missing area due to correction) occurs between the (i-1) th line and the i-th line.

On the other hand, when division control is performed, even if a large offset value 702 is detected, the division control circuit 75 divides the offset value 702 line by line, so that image correlation detection is performed on (i-1) line. The division position correction amount 705 gradually increases or decreases between the eye and the next i-th line.
For this reason, as shown in FIG. 16, the hi-th line set by the division control is set between the (i-1) -th line and the i-th line in the case of "no division control", and (i-
1) It is possible to suppress the occurrence of a region (a mapping missing region due to correction) in which data is not stored between the line of the 1st line and the line of the i-th line. Here, in the hi line, both end pixels are set to P
It is described as s _hi and Pe _hi .

Further, when the correlation target range is widened, the scanning position is corrected by a large offset value, so that there are many areas where data is not stored (map missing areas due to correction). In this case, by performing the division control, it is possible to suppress the generation of a region where the data is not stored in the image memory 6 (a mapping missing region due to the correction), and improve the positional deviation detection accuracy when performing the image correlation detection. As a result, the quality of the image reproduced on the image memory 6 is improved.

Further, the number of pixels is increased by densifying the input image data, and then only the pixels having a small coordinate error at the time of mapping are mapped to the image memory, and the division control is performed. The number of unmapped pixels in memory 6 is greatly reduced. Further, the distortion of the read reproduction image caused by the coordinate conversion error at the time of mapping is significantly reduced. In addition, since it is not necessary to increase the number of pixels of the line image sensor 1 or decrease the maximum manual scanning speed, the image quality on the image memory 6 is further improved.

As described above, according to the configuration of the image processing apparatus of this embodiment, the correction value for correcting the positional deviation is variably controlled at a predetermined rate, and the scanning position is changed based on the stepwise changed correction value. Can be corrected and the image data can be stored in the image memory 6 based on the corrected scanning position data. Therefore, it is possible to suppress the occurrence of a region in the image memory 6 in which the image data is not stored, which is caused by the correction of the scanning position. It Further, since the scanning position is not rapidly corrected, it is possible to reduce the detection range of the correlation value when detecting the positional deviation from the image information between the image data and the stored data,
The circuit can be downsized. Further, since it is possible to prevent an area (image missing) in the image memory 6 where the image data is not stored, it is possible to detect a position shift from the image information between the image data and the stored data in the overlapping scanning areas. Since the error of the correlation value can be reduced, the accuracy of positional deviation detection is improved. As a result, the quality of the image reproduced in the image memory 6 is improved. Further, if the image processing apparatus of the present invention is configured by hardware, it is possible to perform processing sequentially for each line, so that the circuit can be downsized and further processing can be performed in real time.

In this embodiment, the high density image data 500 obtained by densifying the input image data 400 is used to increase the number of pixels, and only the pixels having a small coordinate error during mapping are stored in the image memory. However, the image data 400 that is not densified may be used. In this case, by performing the division control, it is possible to suppress the generation of the non-mapped area. In addition, the determination of the coordinate error at the time of mapping is 0.
In the case of setting 5 (that is, when the coordinate error at the time of mapping is rounded off), R _{i, j} generated between scanning lines in the input image data 400 (P _{i, j} ) and the densified image data 500 is It may be stored at all times (see FIG. 9) and mapped to the image memory 6. As a result, even if a scan line has an unacceptable interval, it is possible to suppress the generation of an area (image missing area) in which image data is not stored.

<Third Embodiment> As described above, in the first and second embodiments, the positional deviation detection of the scanning position in the X and Y directions has been described.
A case in which misregistration detection of the scanning position in the rotation direction is further added will be described with reference to FIGS. 16, 17, and 18.

In the structure shown in FIG. 3, the factors that cause the position error include slippage of the wheels 31 and 32 on the document, accuracy error between the diameters of the wheels 31 and 32, and the distance D between the wheels 31 and 32. Accuracy error, sinking of the wheels 31, 32 into the original, and the center of rotation when rotating by manual scanning is the wheel 3
It can be mentioned that it is affected by the width between 1 and 32. Therefore, by adding the position error with respect to the rotation direction to the correction target, it is possible to accurately correct the position shift.

The image reading section B shown in FIG. 16 is the same as that in the first embodiment. 16 is different from the first embodiment in that a coordinate rotation conversion circuit 8 is added to detect a positional deviation in the rotation direction, and a delay circuit 9 is provided to improve the accuracy of positional deviation detection. Is added. The delay circuit 9 also has an effect of improving the positional deviation correction accuracy in the first and second embodiments. This effect will be described later.

The image processing apparatus shown in FIG. 16 sequentially inputs the scanning position coordinates 300 obtained on a line-by-line basis and the image data 400 corresponding to the scanning position from the image reading section B. The analog output of the line image sensor 1 is the amplifier 1
Amplified by 02 and converted into digital image data by the A / D converter 103. The digital image data is temporarily stored in the image buffer 4. The image buffer 4 outputs the corresponding image data 400 in synchronization with the scanning position coordinate 300.

The correction amount calculation circuit 73 is the image correlation circuit 72.
The offset value 702 output from is accumulated and held. The position correction circuit 74 outputs the position correction amount 70 output from the correction amount calculation circuit 73 that accumulates and holds the position error.
The positional deviation of the scanning position coordinate 300 is corrected by 3 and output as the corrected scanning position coordinate 710.

The delay circuit 9 delays the corrected scanning position coordinates 710 and the image data 400 of the scanning line corresponding to the corrected scanning position coordinates 710 in units of scanning lines. The mapping circuit 5 calculates the pixel data 61 (Pn) of one scanning line from the delayed coordinate data 92 of the corrected scanning position coordinates 710.
Of the delayed image data 91 is stored in the plane image data storage area (memory 65) of the image memory 6, and the pixel data 61 (Pn) to be mapped is stored.

Similar to the first embodiment, if the image memory 6 is constructed as shown in FIG. 7, the storage bit (bit 7) of the write flag for holding the scanning confirmation information of the image data is the memory 66. The storage bits (bits 0 to 6) of the image data correspond to the memory 65. Therefore,
The overlap area detection circuit 71 determines an overlap area to be redundantly scanned from the scan confirmation information stored in the memory address 62 (ADRn) of the memory 66, and outputs the overlap scan area determination signal 701.

The judgment signal 701 checks the write flag (bit 7) of the pixel Pn, and the pixel data 61
It is determined whether the image data has been stored in the address 62 (ADRn) of (Pn). When the bit 7 of the pixel data 61 (Pn) is 1, it indicates that the image data has already been stored in the memory address 62 (ADRn) by the reading scan of the line image sensor 1. Pn) is determined to be included in the overlap scanning area. When bit 7 of the pixel data 61 (Pn) is 0, it is determined that the pixel data 61 (Pn) is included in the new scanning area. Overlap area detection circuit 7
1 outputs the determination signal 701 to the image correlation circuit 72 and the mapping circuit 5
Output to. Here, the determination signal 701 is the pixel data 6
The signal is “0” when 1 (Pn) is included in the new scanning area, and is “1” when included in the overlapping scanning area.

The mapping circuit 5 stores the pixel data 61 (Pn) in the memory 65 when the determination signal 701 is "0". Further, the mapping circuit 5 does not store the pixel data 61 (Pn) in the memory 65 when the determination signal 701 is “1”. This series of processing operations for each pixel is performed for all pixels of the image data 91 of one scanning line.

The coordinate rotation conversion circuit 8 outputs the correlation coordinate 800 in the tilt direction in which the small angle Δφ is swung to the corrected scanning position coordinate 710 [Ps-Pe] to the image correlation circuit 72 in units of one scanning line. .

The image correlation circuit 72 uses the coordinate data of the correlation coordinates 800 to determine the target pixel P for each correlation detection of one scanning line.
The correlation detection address 63 of n is generated, the stored data 64 is read from the storage area (memory 65) of the plane image data of the image memory 6, and at the same time, the scanning confirmation information corresponding to the stored data 64 is read from the memory 66. It is output to the overlap area detection circuit 71.

The overlap area detection circuit 71 determines an overlap area to be scanned redundantly from the scanning confirmation information of the memory 66, and outputs an overlap scanning area determination signal 701. The image correlation circuit 72 performs the correlation value calculation process for the pixel Pn when the determination signal 701 is “1”, and the determination signal 701 is output.
Is 0, the correlation value calculation process for the pixel Pn is not performed.

As shown in FIG. 17, in the image correlation circuit 72, correlation coordinates [Ps-Pe (±
[Delta] [phi])] or correlation coordinates [Ps (± [Delta] [phi])-Pe] is added to perform the correlation detection processing, whereby the positional deviation in the rotational direction is detected, and more accurate positional deviation correction is realized. In this case, the correlation table value h of the first embodiment
When combined with (m, n), the correlation table value is h
(L, m, n). Here, l has a value of -1, 0, 1. 3 ways in the angular direction (rotational direction),
A total of 27 types of correlation tables are created in the position direction (X and Y directions). The minimum value in the correlation table is h (l,
m, n), the angle correction amount Δφoffset _i is 1 · Δφ
It is generated by performing the calculation. Where △
φ is 0.2 degrees, for example. It should be noted that the values of l, m, and n are not specified and are set according to the required positional deviation correction accuracy. If the values of l, m, and n are increased, the area where the correlation is performed becomes wider, and the correction accuracy is improved.

As shown in FIG. 16, in order to detect the rotational direction of the positional deviation, the coordinate of the corrected scanning position coordinate 710 is rotationally converted on the coordinate by the coordinate rotation conversion circuit 8, and the scanning position coordinate 300 [Ps-Pe]. ], The correlation coordinate 800 [Ps-Pe (± Δφ)] or the correlation coordinate 800 [Ps (± Δφ) -Pe] in the tilt direction is output.

The image correlation circuit 72 determines the correlation coordinates 800 in the tilt direction in which the minute angle Δφ is swung by rotation conversion.
According to the above, the above-described calculations of (Equation 10) to (Equation 12) are repeated to create 27 correlation tables 67. Here, l, m, and n take the values of -1, 0, and 1, respectively.

From this correlation table 67, an offset value 702 for correcting the positional deviation is detected. The image correlation circuit 72 holds (l _min , m _min , n _min ) the minimum value of h (l, m, n) when the calculation of the correlation value for each angle of one scanning line is completed. Search and offset value 7
02 to the correction amount calculation circuit 73. There are a plurality of minimum values in the correlation table 67, and the minimum value is (l
_{When min} , _mmin , _nmin ) = (0,0,0) is included, the minimum value of (0,0,0) is preferentially used. Correlation values h (l _min , _mmin ,
The smallest n _min ) means that (Δφ × l _min ,
Image information of the stored data in the image memory 6 and the image data of the line to be mapped from now on when a small value of (Δhx × m _min , Δhy × n _min ) is added to the coordinates of each pixel for mapping. Indicates the best match. Further, when there are a plurality of minimum values and the minimum value includes the center of the correlation window, the offset value 702 in the X direction and Y direction.
The direction is 0. For example, if a 3 × 3 correlation window is set, h (l, 0,0) becomes the center of the correlation window.

The correction amount calculation circuit 73 uses the offset value 70
2 (l _min , m _min , n _min ) is used to calculate the following (Equation 14)
The calculation shown in is performed. Here, the offset value 702 is
X = Δhx × m _min , ΔY = hy × n _min , ΔΨ = Δφ
Xl _min .

[0131]

[Equation 14]

In the above (Equation 14), the suffix i
Represents a position correction amount 703 when the correlation table of the i-th line of the image data 400 is completed. Position correction circuit 7
4 is the scanning position coordinate 300 ((ΔXoffset _i , ΔYoffs
The scanning position coordinate 300 is corrected by adding et _i , Δφoffset _i ) and output as the corrected scanning position coordinate 710.

The scanning position coordinate 300 of the i-th line input to the position correction circuit 74 is set to Ps (Xs, Ys), Pe (X
e, Ye), and the position correction amount 703 is ΔX offset _i , Δ
Yoffset _i, and △ φoffset _i. Position correction circuit 74
Using the scanning position coordinate 300, Psh (Xsh, Ys of the corrected scanning position coordinate 710 is calculated by the following (Equation 15).
h) and Peh (Xeh, Yeh) are calculated.

[0134]

(Equation 15)

Here, the center of rotation on the coordinate is always the coordinate Ps.
And Also, arcsin [] is a function for calculating the arc sine. Nd is the number of read pixels of one line of the line image sensor 1. The center of rotation on the coordinates may be the coordinates Pe. Further, both the rotation conversion with the coordinate rotation center as the coordinate Ps and the rotation conversion with the coordinate rotation center as the coordinate Pe may be performed.
By performing both, it is possible to detect the positional deviation corresponding to both wheels 31, 32.

Next, a method of improving the positional deviation correction accuracy by the delay circuit 9 will be described with reference to FIGS.
This will be described using 8. FIG. 18 is an operation explanatory diagram of the delay circuit shown in FIG.

The division position correction amount (with division control) 705 shown in FIG. 14 is different from the position correction amount (without division control) 703, and since the correction value corresponding to the offset value 702 changes stepwise, the image correlation The necessary correction amount becomes insufficient in the scanning area corresponding to the number of scanning lines during the variable control from the detection position. When the correction amount is insufficient in this way, the correction value for canceling the position error is insufficient, so that connection deviation remains at the position deviation detection position.

In order to avoid such a shortage of the correction amount, the correction scanning position coordinate 71 in the delay circuit 9 is used.
0 and the image data 400 are delayed by an amount corresponding to the number of scanning lines while the division position correction amount 705 is variably controlled, and the mapping circuit 5 and the image correlation circuit 72 are delayed by the mapping circuit 5.
The detection position of the positional deviation detected by is different.

In FIG. 14, since the position where the positional deviation is detected and the position where the division position correction amount 705 is the same as the position correction amount 703 are shifted by four lines, the delay circuit 9 causes the correction scanning position coordinate 710 and the image. The data 400 is delayed by four lines. Due to this delay, at the mapping position in the mapping circuit 5, the same division position correction amount 705 as the position correction amount 703 is obtained. As a result, a correction value for canceling the position error is secured, and the positional deviation correction accuracy is further improved.

Further, as shown in FIG. 18, in the reading area of the line image sensor 1, three states of area a, area b, and a mixture of area a and area b occur at the position of manual scanning. . In the new scan area, it is always area b
In the area where the overlapping scanning area and the new scanning area are mixed, the area a and the area b are mixed, and only the overlapping scanning area is the area a.

Here, in the area a, the memory 65 (see FIG. 1) is used at the position in the correlation detection window when the positional deviation is detected.
The storage state of the planar image data of 6) is the state 1. In this state 1, since the stored data is all around the pixel position of interest, it is possible to detect the correlation value for detecting the displacement.

On the other hand, in the area b, the storage state of the plane image data of the memory 65 is the state 2 at the position in the correlation detection window when the displacement detection is performed. In this state 2, a position where the stored data exists and a position where the stored data does not exist around the pixel position of interest. Therefore, when the positional deviation is detected in the correlation detection window, the correlation values are biased. Because at the position where the stored data exists, the correlation value can be generated from the plane image data and the image information of the stored data, but at the position where the stored data does not exist, there is no stored data, so it is initialized. This is because the correlation value is generated from the data and the plane image data. If the positional deviation is detected from the correlation values generated based on the different data, a detection error will occur.

The scan confirmation information in the memory 66 is used to eliminate the detection error. When it is determined by the scan confirmation information that there is an unstored pixel at any one position in the correlation detection window (that is, in the state 2), the correlation value is not generated. The determination as to whether or not there are unstored pixels is performed using the determination signal 701 detected by the overlapping area detection circuit 71. As a result, the deviation of the correlation value can be eliminated, and the accuracy of the positional deviation detection is improved.

By using the delay circuit 9 shown in FIG. 16, the mapping position and the displacement detection position are made different, and the displacement detection position is preceded with respect to the moving direction of the line image sensor 1. Thus, it is possible to improve the accuracy of positional deviation correction.

By the way, in real time processing,
When the mapping process and the misregistration detection are performed in parallel, pixels newly mapped and stored around the pixel of interest frequently occur, so the correlation value is affected by the new mapped pixel. Further, if the detection position moves back and forth due to camera shake during manual scanning, the influence of the new mapped pixel becomes greater, and the detection accuracy of the positional deviation is reduced. Therefore, by making the mapping position different from the misregistration detection position and allowing the misregistration detection position to precede the moving direction of the line image sensor 1, it is possible to suppress the influence of the new mapping pixel and to detect the misregistration accuracy. Can be improved.

As described above, according to the configuration of the image processing apparatus of the present embodiment, there are such problems as wheel slip, wheel diameter accuracy error, wheel distance accuracy error, and wheel sinking into a document. By adding the correction of the positional deviation with respect to the rotation direction caused by the factor, the position error of the scanning position coordinate 300 is accurately corrected.

Further, the delay circuit 9 delays the corrected scanning position coordinates 710 and the image data 400 to make the mapping position different from the position shift detection position, so that a correction value for canceling the position error at the mapping position is obtained. Since this is ensured, the positional deviation correction accuracy is further improved. Also,
Since it is possible to prevent the data of the line mapped immediately before from affecting the correlation value, the positional deviation correction accuracy is improved.

Incidentally, in the above-mentioned first to third embodiments,
Two encoders 2 for generating a pulse with input image data and scanning position coordinates in accordance with the rotation of two wheels 31, 32 mounted at both ends of the line image sensor 1.
Although the input is made from the hand scanner using a and 2b, the input is not necessarily limited to this configuration. For example, using a tablet or a sheet on which a reference grid is printed as an auxiliary device, the line image sensor 1
Alternatively, the scanning position coordinates and the image data may be input from a hand scanner by detecting the positions of both ends of the scanner (for example, US Pat. No. 4,260,979, US Pat. No. 4,581,761).

In the first to third embodiments, the image processing device for processing the image signal input from the hand scanner using the line image sensor 1 is described as an example. The input device is not necessarily limited to the hand scanner. For example, an image signal from an image input device using an area image sensor may be processed.

The same processing as in the first to third embodiments may be realized by software using a personal computer, MPU (microprocessor unit) or DSP (digital signal processor).

Further, as shown in FIG. 19A, the image processing apparatus of the present invention inputs an image by a small input device 1004 having a limited reading range, and connects and combines partial images on the original 1000. Create a large screen flat image. Therefore, if the image processing apparatus of the present invention is mounted on the main body 1001 of a portable information device, even a portable information device having a small input device 1004 can input an image of a large original. For example, it is most suitable for a mobile FAX, a mobile information terminal, a mobile scanner and the like. Note that FIG.
A display device 1002 includes a liquid crystal panel, a CRT, and the like. Then, the image data read by the input device 1004 is stored in the internal image memory 6, and the image which is connected and combined is displayed on the image memory 6. Further, as shown in FIG. 20, since the input device 1004 can be miniaturized, the input device 1004 can be used integrally with the main body 1001 of the mobile terminal. Further, the input device 1004 shown in FIG. 20 may be configured to be freely attachable / detachable to / from the main body 1001 of the mobile terminal as shown in FIG. The input device 1004 includes a standard I / F (interface)
An IC card 1003 satisfying the specifications is provided,
Data can be transferred and controlled by inserting it into the IC card slot 1007 of the main body 1001. Further, the input device 1004 includes lock pins 1008a, 10
08b is provided, and the lock pin 1008a,
1008b to the pin insertion holes 1009a, 1100 of the main body 1001
The input device 10
04 is fixed to the main body 1001 of the mobile terminal. here,
The input device 1004 is equipped with the line image sensor 1 and wheels 31, 32 shown in FIG. This eliminates the need for auxiliary devices such as tablets, reduces the size,
An input device with excellent operability is realized. If a personal computer is used instead of the main body 1001 of the mobile terminal, a small-sized input device having excellent operability and capable of inputting a large screen can be realized.

Further, as shown in FIG. 19 (e), the image processing apparatus of the present invention performs correction while accumulating the positional deviation of the detected scanning position coordinates, and therefore the input apparatus having a large cumulative position error. Even if 1004 is used, it is possible to correct the positional deviation and generate an image with no connection deviation. Therefore, as used in the first to third embodiments, the two wheels 3 mounted at both ends of the line image sensor 1 are used.
If it is mounted on a hand scanner that uses two encoders 2a and 2b that generate pulses in accordance with rotations of 1 and 32, a greater effect can be obtained. Compared to a hand scanner that uses a tablet or a sheet on which a reference grid is printed as an auxiliary device, it does not require an expensive auxiliary device, so it is possible to realize an image processing device that is compact and has excellent portability and operability. You can Furthermore, it can read thick books. Here, the input device 1004 is, for example, a scanner having the image reading unit B shown in FIG.

[0153]

As described above, according to the image processing apparatus of the present invention, even if the input scanning position includes a position error, the image data and the image data in the overlapping scanning area for overlapping scanning are stored. The positional deviation of the scanning position can be detected from the image information with the data, and the positional deviation can be corrected. Then, by storing the image data in the image memory based on the corrected scanning position, the image connection deviation caused by the position error of the scanning position is prevented, and the image distortion of the planar image connected in the image memory is reduced. can do.

Further, in the configuration of the apparatus of the present invention, correction control means for changing the correction value at a predetermined rate is further provided, and the correction means corrects the scanning position based on the correction value changed at the predetermined rate. According to a preferred example of the above, the correction value for correcting the positional deviation can be variably controlled at a predetermined ratio, and the image data can be stored in the image memory based on the corrected scanning position. Occurrence of a region in the image memory where the image data is not stored is suppressed. In addition, since the scanning position is not rapidly corrected, the detection range of the correlation value when detecting the positional deviation from the image information of the image data and the stored data can be reduced, so that the circuit can be downsized. Furthermore, since the occurrence of an area (image missing) in the image memory where the image data is not stored is prevented, the correlation when detecting the positional deviation from the image information of the image data and the stored data in the overlapping scanning area is prevented. Since the error in the value can be reduced, the accuracy of the positional deviation detection is improved.

Further, in the structure of the apparatus of the present invention, coordinate rotation conversion means for rotating the scanning position at a plurality of predetermined angles about predetermined coordinates is further provided, and the position shift detection means is provided at a plurality of predetermined angles. The coordinate value of each pixel in the image data on the original is calculated based on the coordinates of the scanning position corresponding to the pixel position, and the storage address of the image memory corresponding to the coordinate value is set as the pixel position of interest. Image correlating means for detecting a correlation value between the image data and the stored data stored in the image memory with respect to the peripheral pixel position thereof, and a plurality of the correlations obtained respectively for the predetermined angle. According to a preferable example in which a predetermined angle correction amount preset from the value and a correction amount calculation means for generating a correction value including a predetermined position correction amount are provided, It is possible to detect the positional deviation based on the rotation converted scan position coordinate by the conversion means, it is possible to detect the positional deviation in the rotational direction of the image sensor. As a result, the image quality of the flat image connected in the image memory is improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing an image processing apparatus and an image reading unit for reading and scanning a document image according to a first embodiment of the present invention.

FIG. 2 is an operation explanatory diagram of the position detection circuit according to the first embodiment of the present invention.

FIG. 3 is an explanatory diagram showing calculation of coordinates of pixels read at both ends of the line image sensor according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram of a scanning area of the line image sensor according to the first embodiment of the present invention.

FIG. 5 is a block diagram showing a positional deviation detection circuit according to the first embodiment of the present invention.

FIG. 6 is a block diagram showing a mapping circuit according to the first embodiment of the present invention.

FIG. 7 is an explanatory diagram of an image memory according to the first embodiment of the present invention.

FIG. 8 is an explanatory diagram of a correlation table according to the first embodiment of this invention.

FIG. 9 is an explanatory diagram of pixel density conversion according to the first embodiment of this invention.

FIG. 10 is an address layout diagram of the image memory according to the first embodiment of the present invention.

FIG. 11 is an explanatory diagram of a mapping operation of high density image data onto an image memory according to the first embodiment of the present invention.

FIG. 12 is a block diagram of a positional deviation detection circuit including a division control circuit according to a second embodiment of the present invention.

FIG. 13 is a timing chart of the division control circuit according to the second embodiment of the present invention.

FIG. 14 is an explanatory diagram showing a difference between a position correction amount and a division position correction amount according to the second embodiment of the present invention.

FIG. 15 is an explanatory diagram of a data storage state of the image memory according to the second embodiment of the present invention.

FIG. 16 is a block diagram showing an image processing apparatus and an image reading unit for reading and scanning an original image according to a third embodiment of the present invention.

FIG. 17 is an explanatory diagram of angle correction according to the third embodiment of the present invention.

FIG. 18 is an operation explanatory diagram of the delay circuit according to the third embodiment of the present invention.

FIG. 19 is an explanatory diagram of connection deviation of read images due to positional deviation of scanning positions.

FIG. 20 is a perspective view showing an example of a mobile terminal device according to a third example of the present invention.

FIG. 21 is an exploded perspective view showing another embodiment of the mobile terminal device according to the third embodiment of the present invention.

FIG. 22 is an explanatory diagram showing an example of a mapping missing state in the related art.

FIG. 23 is an explanatory diagram showing another example of a mapping missing state in the related art.

[Explanation of symbols]

 1 Line Image Sensor 2a, 2b Encoder 3 Scanning Position Detection Circuit 4 Image Buffer 5 Mapping Circuit 6 Image Memory 7 Position Deviation Detection Circuit 8 Coordinate Rotation Conversion Circuit 9 Delay Circuit 31, 32 Wheels 33, 35 Position Counter 37 Position Coordinate Detection Circuit 51 Pixel density conversion circuit 52 Coordinate value calculation circuit 53 Integer conversion circuit 54 Address generation circuit 55 Error calculation circuit 65, 66 Memory 67 Correlation table 71 Overlap area detection circuit 72 Image correlation circuit 73 Correction amount calculation circuit 74 Position correction circuit 75 Division control circuit 1001 Mobile terminal body 1004 Input device 1007 IC card slot 1008a, 1008b Lock pin 1009a, 1009b Pin insertion hole

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification number Internal reference number FI Technical indication location G06F 15/66 470 K (72) Inventor ▲ Taka ▼ Hashi Naoki Osaka Prefecture Kadoma 1006 Kadoma 1006 Matsushita Denki Sangyo Co., Ltd.

Claims (27)

[Claims] 1. An image processing apparatus for sequentially inputting image data read by scanning an original image and a scanning position corresponding to the image data, and storing the image data in an image memory based on the scanning position. Therefore, the positional deviation of the scanning position is detected from the image data and the storage data stored in the image memory in the overlapping scanning area for overlapping scanning, and a correction value for correcting the positional deviation is output. At least a position deviation detecting means, a correcting means for correcting the scanning position based on the correction value, and a mapping means for storing the image data in the image memory based on the corrected scanning position are provided. Image processing device. 2. The correction control means for changing the correction value at a predetermined rate is further provided, and the correction means corrects the scanning position based on the correction value changed at the predetermined rate.
The image processing device according to item 1. 3. The correction control means uses a value obtained by dividing the correction value from the positional deviation detection means as a variable control amount, and changes the correction value stepwise at a predetermined rate for each different scanning position. The image processing device according to item 1. 4. The delay means for delaying the corrected scanning position and the corresponding image data is further provided, and the mapping means stores the corresponding image data in the image memory based on the delayed scanning position, The image processing apparatus according to claim 1, wherein a position where the positional deviation of the scanning position is detected is different from a storage position where the image data is stored in the image memory. 5. The misregistration detection means detects the scanning confirmation information of the scanning confirmation information memory and the scanning area holding means for sequentially storing the scanning confirmation information of the image data input to the image memory in the scanning confirmation information memory. The image processing apparatus according to claim 1, further comprising: an overlapping area detecting unit that detects an overlapping scanning area that is overlapped and scanned. 6. The image processing apparatus according to claim 5, wherein the address of the image data stored in the image memory and the address of the scanning confirmation information stored in the scanning confirmation information memory are shared. 7. A memory control means is further provided for controlling the storage of the image data in the image memory according to the determination of the scanning area performed by the overlapping area detection means, wherein the memory control means is an overlapping scanning area that is scanned in an overlapping manner. The image processing apparatus according to claim 5, wherein storage of image data in the image memory is prohibited when it is determined that the image data exists. 8. The misregistration detection means sets a storage address of an image memory corresponding to a coordinate value on the original of each pixel in the image data calculated based on the scanning position as a pixel position of interest, and sets the pixel position of interest. An image correlation unit that detects a correlation value between the image data and the stored data stored in the image memory for each of the peripheral pixel positions that include the correction value and a correction value that is set in advance from the plurality of correlation values. The image processing apparatus according to claim 1, further comprising: 9. The position shift detecting means generates a plurality of correlation values only when all of the peripheral pixel positions including the pixel position of interest in the pixel positions to be correlated are overlapping scanning regions. The image processing device described. 10. The image according to claim 1, wherein the correction unit includes a holding unit that generates a cumulative correction value by accumulating the correction values output from the positional deviation detection unit, and corrects the scanning position by the cumulative correction value. Processing equipment. 11. The mapping means includes an address generating means for generating a storage address of an image memory corresponding to a coordinate value of each pixel in the image data on the document based on the scanning position, and each pixel in the image data. The image processing apparatus according to claim 1, further comprising a data storage unit that stores the data at the storage address of the image memory. 12. The image data and the scanning position are input from an input device including an image sensor for reading an image and a position detecting means for detecting the reading position of the image sensor by two detecting means. The image processing device described. 13. The image processing apparatus according to claim 12, wherein the image sensor is a line sensor. 14. The image processing apparatus according to claim 12, wherein the input device is a hand scanner. 15. The image processing apparatus according to claim 12, wherein the detection means includes a wheel that rotates in contact with the document surface and an encoder that outputs a pulse corresponding to the rotation of the wheel. 16. The image processing apparatus according to claim 15, wherein the encoder outputs pulses of two systems having different phases. 17. The image processing apparatus according to claim 12, wherein the input device is detachably attached to the main body of the portable information device. 18. The input device according to claim 17, wherein the input device is attached to the body of the portable information device by inserting a lock pin provided in the input device into a pin insertion hole provided in the body of the portable information device. Image processing device. 19. The mapping means comprises pixel density conversion means for densifying image data from adjacent pixels by interpolation,
Means for calculating the real number coordinate value on the original of each pixel in the high-density image data based on the scanning position in units of the predetermined pixel density, and the real number coordinate value on the original for the predetermined number. Integer conversion means for converting the pixel density of each unit into an integer coordinate value, means for calculating a coordinate error value generated when the coordinate values are converted into integers, and means for converting the integer coordinate values into an address of an image memory. And a means for comparing the magnitude of the coordinate error value with a predetermined value determined in advance, and only the pixels having the magnitude of the coordinate error value equal to or less than the predetermined value are stored in the address of the image memory. The image processing device according to claim 11. 20. The pixel density conversion means comprises means for generating an interpolated pixel by using a pixel to be processed and adjacent pixels around the pixel to be processed, and the image data is converted to a double pixel density by the means. The image processing device described. 21. A pixel to be processed P _{i, j} (however, this represents the j-th pixel data of the image data of the i-th line in the image data) and its neighboring pixels P _{i, j + 1} , P _{i. -1, j} , P
_{Using i-1, j + 1} , interpolation pixel Q _{i, j}
The image processing apparatus according to claim 20, wherein R _{i, j} and S _{i, j} are calculated. [Equation 1] 22. The image processing apparatus according to claim 19, wherein the predetermined value is 0.3 to 0.35 with a predetermined pixel density as a unit. 23. Coordinate rotation conversion means for rotating the scanning position at a plurality of predetermined angles about a predetermined coordinate is further provided, and the position shift detecting means is adapted to coordinate the scanning position corresponding to the plurality of predetermined angles. Based on this, the coordinate value of each pixel in the image data on the original is calculated, and the storage address of the image memory corresponding to the coordinate value is set as the target pixel position, and the surrounding pixel positions including the target pixel position are calculated. Image correlation means for detecting a correlation value between each of the image data and the stored data stored in the image memory, and a predetermined value preset from a plurality of the correlation values obtained for the predetermined angle. The image processing apparatus according to claim 1, further comprising a correction amount calculation unit that generates a correction value including an angle correction amount and a predetermined position correction amount. 24. The misalignment detection means is configured to perform a plurality of correlations only when all the peripheral pixel positions including a pixel position of interest obtained for each rotation angle are overlapping scanning regions at a pixel position to be correlated. 24. Generating a value
The image processing device according to item 1. 25. The image processing apparatus according to claim 23, wherein the image correlation unit generates a plurality of correlation values obtained for each rotation angle. 26. The image data and the scanning position are input from an input device equipped with a line image sensor for reading an image and a position detecting means for detecting the reading position of the line image sensor by means of a coordinate rotation. 24. The image processing apparatus according to claim 23, wherein the conversion unit rotationally converts the scanning position coordinates of both ends of the line image sensor in a plurality of directions on the coordinates in units of scanning lines. 27. The image data and the scanning position are input from an input device equipped with a line image sensor for reading an image and a position detecting means for detecting the reading position of the line image sensor by means of an image correlation. The image processing apparatus according to claim 23, wherein the means generates a plurality of correlation values obtained for each rotation angle of each scanning line.