JP2000099693A - Image connection system and storage medium recording image connection program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simply and clearly input the positional relations among plural images by collating these images with each other and at the same time inputting the corresponding points of the images on a screen. SOLUTION: When an image connection program is activated on a computer 11, an MPU 12 produces a dialog on a monitor screen. The dialog includes two windows which are placed side by side and display a basic image and its connected image inputted via an image input device 21, etc., respectively. In such a condition where the dialog is displayed, the MPU 12 waits for the mouse input operations of an operator. The operator collates two images included in the dialog with each other and decides at least two corresponding points overlapping each other between both images. For instance, those two images are shown on a map and the operator finds out a crossroads which are common to both map images. Then the operator decides the upper and lower right corners of the crossing part of the crossroads as the corresponding points, puts a mouse pointer on these corresponding points and clicks a mouse.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image connection system for performing connection processing of a plurality of images, and a recording medium storing an image connection program. In particular, the present invention relates to a technology of a man-machine interface for easily and reliably inputting a positional relationship between these images.

[0002]

2. Description of the Related Art Conventionally, an algorithm for performing template matching or structure matching has been known as an algorithm of an image connection system. Among these, the image connection system using template matching executes the image connection processing in the following procedure.

(Step 1) Image data in a partial range is cut out from an image to be connected (hereinafter, referred to as a "connected image") to create a template image. (Step 2) A partial range is initially set within a range of a reference image (hereinafter, referred to as a “base image”). (Step 3) A difference for each pixel is calculated between a partial range of the base image and the template image, and a sum of absolute values of these pixel differences (so-called “residual”) is calculated. If the intermediate result of the residual exceeds a predetermined threshold during this calculation,
It is determined that there is no possibility that the partial range of the base image matches the template image, and the calculation of the residual is terminated at that time.

(Step 4) It is determined whether or not the final result of the residual is equal to or smaller than an allowable value. If it is larger than the allowable value, it is determined that the partial range of the base image does not match the template image. In this case, the operation returns to step 3 after the setting of the partial range is slightly shifted on the base image. On the other hand, if it is equal to or smaller than the allowable value, it is determined that the partial range of the base image matches the template image, and the operation shifts to step S5.

(Step 5) The connected image is translated so that the template image overlaps a part of the base image,
The base image and the connected image are connected. In addition to the algorithm for calculating the residual as described above (residual sequential test method), an algorithm for judging image coincidence from a cross-correlation coefficient between a base image and a template image is known (for details, see 19).
The University of Tokyo Press "Image Analysis Handbook" published in 1991
707-712).

On the other hand, as an image connection system using structured matching, there is known an image connection system that extracts characteristic information such as a branch point from an image and performs image position adjustment based on the characteristic information (detailed description). Is described in the above-mentioned “Image Analysis Handbook”, pp. 713 to 746). Also, recently, a method using FFT transform or Gabor Wavelet transform for extracting feature information is known.

[0007]

In the conventional example using the template matching described above, if there is a difference in the rotation angle or a difference in the scale ratio between the base image and the connected image, the matching judgment is made. In doing so, the rotation angle and scale must be changed many times. For this reason, there has been a problem that the number of calculations of the residual and the like is greatly increased.

On the other hand, in the conventional example using the structured matching described above, an amount invariant to the rotation angle or the like (for example, an FFT transform coefficient calculated on a circle) can be used as the feature information. Therefore, in theory, such an alignment of the feature information makes it possible to obtain the amount of translation of the image without being affected by the difference in the rotation angle.

However, in a digital image, if the rotation angle and the image size are different, the pixel positions between the images do not correspond one-to-one. Therefore, for a low-resolution digital image, there is a problem that the invariance of feature information is not so established, and it is difficult to apply structured matching.

In addition, the following (a) to (a) are common to the conventional examples of the template matching and the structured matching described above.
There is also a problem that the image as shown in FIG. (A) The case where the gradation change of the image is small or gradual (b) The case where the gradation change of the image is extremely large or periodic (c) The overlapping area of the image is small, and Therefore, in order to solve the above-described problems, a man-machine interface for simply and surely inputting a positional relationship between images is provided in the inventions according to claims 1 to 8. It is an object of the present invention to provide an image linking system provided with the same.

[0012] In particular, the invention according to claim 2 provides an image connection system capable of reliably executing image connection processing even if an input corresponding point (described later) has an error. Aim. It is another object of the present invention to provide an image connection system capable of reliably executing image connection processing even when the scale ratio between images is different.

An object of the present invention is to provide an image connecting system capable of cutting unnecessary portions of each image generated at the time of connecting images by simple processing. An object of the present invention is to provide an image connecting system capable of cutting unnecessary portions of each image in an appropriate direction. An object of the present invention is to provide an image connection system capable of reducing the disappearance of image information after connection processing.

An object of the present invention is to provide an image connection system capable of increasing the speed of scroll display when performing scroll display while changing the display angle of an image. It is an object of the present invention to provide a specific form of an image connection system to which an object-oriented database is added.

According to the tenth aspect of the present invention, in the first aspect,
It is another object of the present invention to provide a recording medium on which an image connection program for realizing the image connection system according to Item 6 on a computer is recorded. In the invention according to claim 11,
It is an object of the present invention to provide a recording medium storing an image connection program for realizing the image connection system according to claims 7 to 9 on a computer.

[0015]

According to a first aspect of the present invention, there is provided an image connecting system for connecting a plurality of images, wherein the plurality of images are displayed together or sequentially on a screen. And a position input unit for receiving at least two corresponding points overlapping among a plurality of images, and a “corresponding point of one image” obtained by the position input unit and a “corresponding point of another image”. Using a translation amount calculation unit that calculates the translation amount and rotation amount for overlapping, and using the translation amount and rotation amount calculated by the translation amount calculation unit, other images are coordinate-transformed, and are displayed on a common coordinate space. An image arrangement unit for arranging a plurality of images.

In such a configuration, a position input unit is newly provided as a man-machine interface. as a result,
The operator can find the corresponding points overlapping between the images while collating the plurality of images on the screen, and can perform an input operation easily and quickly. On the other hand, the movement amount calculation unit calculates the amount of translation and rotation of the image from at least two corresponding points acquired from the position input unit. The image arranging unit executes the coordinate transformation of the image using the translation amount and the rotation amount. Therefore, appropriate connection processing can be performed on images having different rotation angles.

According to a second aspect of the present invention, in the image linking system according to the first aspect, the movement amount calculating section includes one corresponding point and another corresponding point in each image. And a rotation amount calculation unit that calculates a relative angle of the vector between a plurality of images and obtains a rotation amount from the relative angle, and calculates a distance between one corresponding point between the plurality of images. And a translation amount calculating unit for calculating the translation amount from the distance.

Normally, the corresponding points from the position input unit are input by the operator, and therefore naturally include some errors. For this reason, it is expected that the translation amount and the rotation amount cannot be calculated under strict conditions such that all the corresponding points are completely overlapped between the images. However,
According to the second aspect of the present invention, a vector which connects one corresponding point and another corresponding point is determined, and first, a rotation amount is obtained from a relative angle between these vectors. continue,
The amount of parallel movement is obtained from the distance of one corresponding point.

Such translation and rotation amounts are numerical values calculated under the following conditions. (1) One corresponding point at least overlaps. (2) The slopes of the vectors connecting one corresponding point and the other corresponding points match. Since such mild conditions are set, it is possible to always calculate the amount of translation and the amount of rotation even if the corresponding points include some errors, and it is possible to reliably execute the image connection processing. .

According to a third aspect of the present invention, in the image connection system according to the first or second aspect, the movement amount calculation unit calculates an interval between corresponding points in each image, A scale calculating unit for calculating a scale relationship between the plurality of images based on the ratio of the intervals; the image arranging unit performs enlargement or reduction of the image based on the scale relationship obtained by the scale calculator, and arranges the images on a common coordinate space; It is characterized in that the scale relationships of the images to be processed are matched.

With such a configuration, it is possible to perform an accurate connection process even on a plurality of images having different scale ratios. In particular, in this configuration, the corresponding points acquired from the position input unit are diverted to determine the scale relationship between the images. Therefore, the operator does not need to perform any troublesome work such as calculating and inputting the scale ratio one by one.

(Claim 4) According to a fourth aspect of the present invention, in the image connection system according to any one of the first to third aspects, the image arranging section stores each image on a common coordinate space. When arranging, unnecessary portions of each image are cut by line segments passing through the corresponding points obtained in the position input unit.

By the way, regarding the processing of the overlapping area caused by the image concatenation, the processing of calculating the average value and the intermediate value (the case where three or more images overlap) between the overlapping pixels and the processing of the image (mainly the overlapping area) are performed. A process of dividing and cutting one as an unnecessary part can be considered. In particular, in the case of the latter processing, it is necessary to appropriately set a cut line that reliably divides the overlapping area, and usually, complicated calculations and determination processing are indispensable.

However, in the present invention, a line segment passing through the corresponding point is simply used as the cut line. Generally, the corresponding points are guaranteed to be located in the overlapping area between the images. Therefore, the line segment passing through the corresponding point always divides the overlapping area, and is a setting that is approximately appropriate as a cut line. Therefore, it is possible to set a cut line very easily and cut unnecessary portions of each image without much complicated calculation and determination processing.

Further, the operator can control the setting of the cut line to some extent depending on the position of the corresponding point to be given. Therefore, it is possible to realize an excellent man-machine interface that can reflect the intention of the operator on the setting of the cut line.

The "line segment" described in claim 4 is:
The invention is not limited to a line segment passing through only two corresponding points. For example, a line segment having an appropriate inclination passing through one corresponding point or a line segment passing through three or more corresponding points may be used. The “line segment” described in claims 4 to 6 is not limited to a straight line. It may be a bent line or a curve (for example, a Bezier curve, a multi-degree curve, a hyperbola, etc.).

According to a fifth aspect of the present invention, in the image connection system according to the fourth aspect, the image arranging section connects the images from a plurality of line segments passing through two or more corresponding points. It is characterized in that a line segment whose angle with the direction is close to a right angle is selected, and unnecessary portions of each image are cut using the selected line segment.

In the above-described configuration, a line segment that is as orthogonal as possible to the connecting direction of images is selected and set as a cut line. By setting such a cut line, an image (mainly, an overlapping area) is divided in a connecting direction, and one of the divided areas is cut as an unnecessary part. As a result, the connected portions of the plurality of images are cut in an appropriate direction. (Claim 6) In the image connection system according to claim 4, in the image connection system according to claim 4, the image arranging unit determines an unnecessary portion to be cut from a plurality of line segments passing through two or more corresponding points. A line segment having the smallest total area is selected, and unnecessary portions of each image are cut using the selected line segment.

In general, when only an overlapping area is divided by a cut line and one of the overlapping areas is cut as an unnecessary part, the image information after the concatenation processing does not disappear at all. However, cut the entire image at the cut line,
If one of them is cut as an unnecessary part, image information other than the overlapping area may be cut as an unnecessary part.

Therefore, in the above-described configuration, a line segment in which the total area of unnecessary portions to be cut is as small as possible is selected as a cut line. By setting such a cut line,
It is possible to suppress the disappearance of the image information after the connection processing. (Claim 7) According to the invention described in claim 7, in the image connection system according to any one of claims 1 to 6, a plurality of images arranged on a common coordinate space are externally input. The image storage unit that performs rotation processing according to the image display angle to be stored and stores as a connected image, and the connected images stored in the image storage unit are sequentially displayed according to the scroll direction, and scrolls over a plurality of connected images. And a scroll display unit for performing display.

In such a configuration, since the image rotation processing according to the image display angle is completed during image connection, the image rotation processing during scroll display can be omitted. Therefore, the scroll display speed can be further increased. (Claim 8) The invention according to claim 8 is the image connection system according to claim 7,
The image storage unit is provided from the outside with an object-oriented database in which an object including at least one of moving image information, still image information, audio information, command information, or character information can be registered in association with the stored connected image. An object execution unit that selects an object from an object-oriented database in accordance with a position indication on the screen, and executes the selected object at the timing of image display.

According to a ninth aspect of the present invention, in the image linking system according to the seventh aspect, the image storage section associates the moving image information, the still image information, Executes at least one of the following (1) and (2) in correspondence with an object-oriented database in which an object including at least one of voice information, command information and character information can be registered, and a position designation on a screen given from outside And an object updating unit for updating the object-oriented database.

(1) A process of selecting a registered object corresponding to the position indication from the object-oriented database and executing one of editing, updating, and deleting for the selected registered object. (2) The position A process for newly registering an object in the object-oriented database in association with an instruction. (Claim 10) The recording medium according to claim 10, wherein the recording medium is a computer, and the recording medium according to any one of claims 1 to 6. An image connection program for functioning as a position input unit, a movement amount calculation unit, and an image arrangement unit is recorded.

By executing such an image connection program on a computer, the image connection system according to any one of claims 1 to 6 can be realized. (11) A recording medium according to (11), comprising: a computer configured to execute a position input unit, a movement amount calculation unit, an image arrangement unit, an image storage unit, An image connection program for functioning as a scroll display unit is recorded.

By executing such an image connection program on a computer, an image connection system according to any one of claims 7 to 9 can be realized.

[0036]

Embodiments of the present invention will be described below with reference to the drawings.

<First Embodiment> The first embodiment is an embodiment corresponding to the first, second, third, fourth, fifth and tenth aspects of the present invention.

FIG. 1 is a diagram showing the overall configuration of an image connection system using a computer 11. In FIG. 1, an MPU (microprocessor) 12 is provided inside a computer 11. An input device 13 including a keyboard and a mouse, a hard disk 16, a memory 17, an image processing board 18, and an interface board 20 are connected to the MPU 12. A monitor 19 is connected to an image output terminal of the image processing board 18. On the other hand, an image input device 21 such as a scanner or an electronic camera is connected to the interface board 20.

On the other hand, a CD-ROM drive device 22 is connected to the MPU 12. Into the CD-ROM drive 22, a CD-ROM 23 in which an image connection program and its installation program are recorded is inserted. With the installation program in the CD-ROM 23, the MPU 12 develops the image connection program in the CD-ROM 23 and stores it in the hard disk 16 in an executable state.

(Correspondence between the present invention and the first embodiment)
Hereinafter, the correspondence relationship with the configuration requirements of the present invention will be described. First, regarding the correspondence between the invention described in claim 1 and the first embodiment, the position input unit is connected to the input device 13 and the MP.
The movement amount calculation unit corresponds to “the function of displaying the base image and the connected image and obtaining the mouse input of the corresponding point” of U12.
The image placement unit corresponds to “the function of calculating the rotation amount and the translation amount of the connected image based on two or more corresponding points” of the PU 12, and the MPU 12 “coordinates the connected image according to the rotation amount and the translation amount” Function to convert the base image and the connected image on the same address space of the memory 17 ".

In addition, regarding the correspondence between the invention described in claim 2 and the first embodiment, the rotation amount calculation unit is configured to control the MPU 1
2 corresponds to “the function of calculating the amount of rotation based on the relative angle of a vector connecting two corresponding points”, and the parallel movement amount calculating unit calculates the parallel movement amount of the MPU 12 from the distance between the one corresponding point. Function to calculate ". Further, claim 3
In the correspondence between the invention described in (1) and the first embodiment, the scale calculation unit corresponds to the “function of calculating the scale ratio between the base image and the connected image from the interval ratio between two corresponding points” of the MPU 12. The image arranging section corresponds to the “function of arranging the base image and the connected image in the same address space of the memory 17 after scaling the connected image and adjusting the scale ratio to the base image” of the MPU 12.

Further, regarding the correspondence between the invention described in claim 4 and the first embodiment, the image arrangement unit is configured to use the MPU 12
"The function of arranging the base image and the connected image on the same address space of the memory 17 after removing the unnecessary portion of each image by the line segment passing through the corresponding point". Furthermore, regarding the correspondence between the invention described in claim 5 and the first embodiment,
The image arranging unit is configured by the MPU 12 “in the line passing through the corresponding point,
A function of removing unnecessary portions of each image with a line segment that is as orthogonal as possible to the connecting direction ”.

Further, as for the correspondence between the invention described in claim 10 and the first embodiment, the recording medium is a CD-RO.
Corresponds to M23. (Operation of First Embodiment) The operation of the first embodiment will be described below. FIG. 2 is a flowchart illustrating the operation of the first embodiment.

First, when the image connection program is started in the computer 11, the MPU 12 creates a dialog on the monitor screen. FIG. 3 is a diagram showing a display example of such a dialog. In FIG. 3, two windows are arranged in a dialog. In these windows, the base image and the connected image input from the image input device 21 and the like are displayed, respectively. With such a dialog displayed, the MPU 12 waits for a mouse input by the operator (S1 in FIG. 2).

The operator compares the two images on the dialog and determines at least two overlapping corresponding points between the two images. For example, the two images displayed in FIG. 3 are map images. The operator finds a crossroad common to these map images, and determines the upper right corner and the lower right corner of the intersection of the crossroads as the corresponding points A and B, respectively. The operator determines the corresponding points A and B of the two images determined in this way.
The mouse pointer over and then perform a mouse click.

The MPU 12 sequentially takes in these mouse click positions via the input device 13 and stores them in the following variables, respectively. (1) Relative coordinates of the corresponding point A on the base image (OLD # XA, OL
D # YA) (2) Relative coordinates of corresponding point A on the connected image (NEW # XA,
NEW # YA) (3) Relative coordinates of the corresponding point B on the base image (OLD # XB, OL
D # YB) (4) Relative coordinates of corresponding point B on the connected image (NEW # XB,
NEW # YB) Note that by making these variable groups into an expandable array structure or list structure, it becomes possible to receive input of three or more corresponding points by the operator.

FIG. 4A is a diagram showing the corresponding points A and B input as described above. The MPU 12 calculates, for each coordinate system of the two images, the gradients OLD # γ and NEW # γ of the vector AB connecting the corresponding points A and B as follows. OLD # γ = tan ^-1 [(OLD # YB-OLD # YA) / (OLD # XB-OLD # XA)] + π ... [Equation 1] NEW # γ = tan ^-1 [(NEW # YB- NEW # YA) / (NEW # XB-NEW # XA)] + π (Expression 2) The MPU 12 calculates the relative angle (OLD # γ-NEW # γ) of the vector AB from the two slopes obtained in this manner. ) Is calculated and set as the rotation amount NEW # θ of the connected image (S2 in FIG. 2).

Next, the MPU 12 determines the corresponding point A on the base image.
Then, the parallel movement amount P (OLD # XA-NEW # XA, OLD # YA-NEW # YA) of the corresponding point A is calculated from the distance between the corresponding point A and the corresponding point A on the connected image (S3 in FIG. 2).

Further, the MPU 12 calculates the scale ratio S from the ratio between the magnitudes of the vectors AB in the two images as follows (S4 in FIG. 2). S = √ [[(OLD # YB-OLD # YA) ² + (OLD # XB-OLD # XA) ² ] / [(NEW # YB-NEW # YA) ² + (NEW # XB-NEW #XA) ² ]] [Equation 3] With a series of calculations as described above, all parameters necessary for coordinate conversion of the connected image are prepared.

Here, the MPU 12 determines whether there are three or more corresponding points (S5 in FIG. 2). If there are only two corresponding points, the MPU 12 shifts the operation to step S7 in order to set a cut line from these two corresponding points. On the other hand, when there are three or more corresponding points, some line segments passing through these corresponding points can be considered. Therefore, MPU12
Determines a plurality of line segments passing through two or more corresponding points, and selects a line segment that is as orthogonal as possible to the connecting direction of images as a cut line (S6 in FIG. 2).

Using the cut line thus selected, the MPU 12 divides the overlapping area between the images, and removes one of the overlapping areas from the memory 17 as an unnecessary part (FIG. 2S).
7). FIG. 5 is a diagram for specifically explaining such an unnecessary portion cutting process. In FIG. 5A, three corresponding points A to C are input from the operator side. First, the MPU 12 selects 2 out of these corresponding points A to C.
Three types of line segments AB, BC, and CA connecting the corresponding points of the points are determined. Next, the MPU 12 selects, from among these line segments,
A line segment AB orthogonal to the image connecting direction (here, the left and right direction) is selected as much as possible and set as a cut line.

As shown in FIG. 5B, the MPU 12
Using this cut line, the overlapping area of each image is divided, and the different ones are stored in the memory 17 as unnecessary parts.
Remove from above. By removing such unnecessary portions, as shown in FIG. 5C, the connected portions of the image are appropriately processed without any loss of the connected image information.

Subsequently, the MPU 12 sequentially executes coordinate conversion of the connected image. First, as shown in FIG. 4B, the MPU 12 rotates the connected image around the corresponding point A by the amount of rotation.
Rotate by NEW # θ (S8 in FIG. 2). Then, MPU1
2 reduces (enlarges) the connected image by the scale ratio S around the corresponding point A, as shown in FIG.
9).

Further, as shown in FIG. 4D, the MPU 12 translates the connected image by the parallel movement amount P, and arranges the base image and the connected image in the same address space of the memory 17 (FIG. 4D). 2S10). Note that these steps S8
If the pixel positions before and after the conversion do not correspond one-to-one in the coordinate conversion of to 画素, the pixel is resampled. Further, in order to reduce the deterioration of the connected image due to such resampling of pixels, the size of the connected image may be enlarged prior to the coordinate transformation.

(Effects of First Embodiment, etc.) According to the above-described series of processing, in the first embodiment, two images can be connected in a correct positional relationship. In particular, since the operator can check a plurality of images on the monitor screen, the operator can easily and quickly find the corresponding points overlapping between the images and immediately input them via the input device 13 such as a mouse. Can be.

On the other hand, on the image connection system side, by acquiring at least two corresponding points from the input device 13, the parallel movement amount P and the rotation amount NEW # θ of the image can be calculated. Therefore, it is possible to perform accurate connection processing even for images having different rotation angles. In particular,
In the process of such concatenation processing, there is no need to perform complicated residual calculation and feature information extraction processing as in the related art. Therefore, the processing time of the image connection can be significantly reduced as compared with the conventional example. Further, even for an image of a type for which conventional template matching or structure matching is not suitable, image connection can be performed without any problem.

Furthermore, in the first embodiment, the gradual “the one corresponding point A at least overlaps” and the “gradient of the vector connecting the one corresponding point A and the other corresponding point B match” are described. Under the conditions, the translation amount P and the rotation amount NEW # θ are calculated. Accordingly, even if the positions of the corresponding points A and B include some error, the image connection processing can be performed flexibly and reliably.

In the first embodiment, the scale ratio S between images is calculated based on the ratio of the intervals between the corresponding points AB.
Therefore, it is possible to perform an accurate connection process on a plurality of images having different scale ratios. In particular, since the scale ratio between the images is obtained by diverting the input information of the corresponding points, there is no need for the operator to calculate the scale ratio one by one and input it separately.

Further, in the first embodiment, since a line segment passing through the corresponding point is used as the cut line, complicated processing for properly determining the cut line is not required. Also,
On the operator side, the setting of the cut line can be controlled to some extent by the position of the corresponding point to be given. In particular, in the first embodiment, since a line segment that is as orthogonal as possible to the connecting direction of the images is selected as the cut line, it is possible to cut the connecting portions of a plurality of images in an appropriate direction.

In the first embodiment described above, the case where two images are connected is described, but the present invention is not limited to this. For example, at least two overlapping corresponding points may be input for each of three or more images. In such a case, three or more images can be connected at a time. In the above-described first embodiment, the cut line is selected from the line segments passing through the two corresponding points.
It is not limited to this. For example, one or more corresponding points may be selected, and a line segment passing through the selected corresponding point may be used as the cut line.

Furthermore, in the first embodiment described above, the case where images are connected in the left-right direction has been described, but the present invention is not limited to this. For example, as shown in FIG. 6, images may be connected in the vertical direction. FIG.
In the case shown in (1), a line segment AB that is as orthogonal as possible to the direction in which the images are connected (in this case, the vertical direction) may be selected as the cut line.

In the above-described first embodiment, since the connection direction of the images is processed as a rough direction such as up and down, left and right, and diagonal, the cut line selection processing can be greatly simplified. it can. In practice, such a simple process is often sufficient. However, it is not necessary to be limited to such processing. For example, the connection direction may be obtained in detail from the inclination of the parallel movement amount P. Also,
Based on the parallel movement amount P and the rotation amount NEW # θ, the “slope of the line connecting the center point of the base image and the center point of the connected image” is calculated on the common coordinate space, and the connection direction is calculated from the slope. You may ask for details.

Further, in the first embodiment described above, (1) rotation about the corresponding point A (2) enlargement / reduction around the corresponding point A (3) parallel movement by the distance of the corresponding point A The coordinate transformation of the connected image is being executed. However, it is a matter of course that a conversion formula for performing the coordinate conversions (1) to (3) at once may be created and the coordinate conversion may be completed at once.

In particular, the parameters (P, NEW # θ, S) used for the coordinate transformations (1) to (3) are calculated from the two corresponding points in an almost shortest efficient calculation procedure. It is possible. Therefore, in the first embodiment, by performing the coordinate conversion procedure (1) to (3), it becomes possible to execute the parameter calculation for the coordinate conversion at a particularly high speed.

However, the procedure of coordinate conversion is (1)
It is not limited only to (3). In general, you can change the order of coordinate transformation, shift the center of rotation transformation,
Equivalent coordinate transformation can be obtained indefinitely by shifting the center of enlargement / reduction transformation or adding invalid coordinate transformation. Therefore, it is of course possible to use any of these equivalent coordinate transformations.

In the first embodiment, the CD-ROM 23 is used as a recording medium for the image connection program. However, the type of the recording medium is not limited to this. In general, any recording medium may be used as long as it is a machine-readable medium. Further, in the above-described first embodiment, the base image and the connected image are simply arranged and displayed together. Such a display method has an advantage that both images can be easily collated, and a corresponding point can be found quickly and easily. However, the present invention is not limited to this display method. For example, the base image and the connected image may be displayed sequentially. In addition, the base image and the connected image may be partially enlarged and displayed according to an instruction from the operator. According to such a partially enlarged display,
It is possible to further improve the input accuracy of the corresponding points.

Next, another embodiment will be described. <Second Embodiment> A second embodiment is described in claims 1, 2, and 3.
This is an embodiment corresponding to the invention described in 4, 6, 7, 8, 9, and 11. The system configuration of the second embodiment is as follows.
Since the configuration is the same as that of the first embodiment (FIG. 1), the duplicate description is omitted here.

(Correspondence between the present invention and the second embodiment)
Hereinafter, the correspondence relationship with the configuration requirements of the present invention will be described. First, regarding the correspondence between the invention described in claim 1 and the second embodiment, the position input unit is configured so that the input device 13
The movement amount calculation unit corresponds to “the function of displaying the base image and the connected image and obtaining the mouse input of the corresponding point” of U12.
The image placement unit corresponds to “the function of calculating the rotation amount and the translation amount of the connected image based on two or more corresponding points” of the PU 12, and the MPU 12 “coordinates the connected image according to the rotation amount and the translation amount” Function to convert the base image and the connected image on the same address space of the memory 17 ".

The correspondence between the second embodiment and the second embodiment is as follows.
2 corresponds to “the function of calculating the amount of rotation based on the relative angle of a vector connecting two corresponding points”, and the parallel movement amount calculating unit calculates the parallel movement amount of the MPU 12 from the distance between the one corresponding point. Function to calculate ". Claim 4
As for the correspondence between the invention described in (1) and the second embodiment, the image placement unit removes the unnecessary portion of each image by a line segment passing through the corresponding point of the MPU 12 and then stores the unnecessary portion on the same address space of the memory 17. Function of arranging images and connected images ”.

The correspondence between the invention described in claim 6 and the second embodiment is as follows.
The “function of removing unnecessary portions of each image using a line segment having the smallest total area of unnecessary portions among the line segments passing through the corresponding points”. Further, the invention described in claim 7 and the second invention
The image storage unit corresponds to the hard disk 16 and the “function of storing the connected image rotated according to the image display angle in the hard disk 16” of the MPU 12, and the scroll display unit corresponds to the MPU 12.
(A function of executing scroll display while sequentially switching a plurality of connected images on the hard disk 16).

As for the correspondence between the invention described in claim 8 and the second embodiment, the object-oriented database corresponds to the “function for registering and managing connected images and objects” of the hard disk 16 and the MPU 12. The object execution unit corresponds to the “function of executing an object in response to a mouse input” of the MPU 12.

Further, regarding the correspondence between the invention described in claim 9 and the second embodiment, the object-oriented database corresponds to the “function for registering and managing connected images and objects” of the hard disk 16 and the MPU 12. The object update unit corresponds to the “function of updating an object in response to a mouse input” of the MPU 12. Regarding the correspondence between the invention described in claim 11 and the second embodiment, the recording medium corresponds to the CD-ROM 23.

(Operation of Second Embodiment) The operation of the second embodiment will be described below. FIGS. 7 and 8 are flowcharts for explaining the operation of the second embodiment. First, the operator starts an image connection program in the computer 11, and selects a base image and a connection image from a plurality of images captured from a scanner or the like (FIG. 7S
1).

Note that the base image here is an image selected first in the connection work or an image whose connection in the common coordinate space is already known after the connection work has been completed. On the other hand, the connected image is an image that partially overlaps the base image. The MPU 12 displays the two images thus selected together on a monitor screen. In such a display state, the MPU 12 waits for a mouse input by the operator.

The operator collates the two images on the monitor screen and determines at least two overlapping points between the two images. The operator places the mouse pointer on the corresponding point of each image determined in this way and executes a mouse click (S2 in FIG. 7). The MPU 12 sequentially takes in the positions of these mouse clicks via the input device 13 and stores them in the following variables.

(1) Relative coordinates of corresponding point A on the base image (OLD # XA, OLD # YA) (2) Relative coordinates of corresponding point A on the connected image (NEW # XA,
NEW # YA) (3) Relative coordinates of the corresponding point B on the base image (OLD # XB, OL
D # YB) (4) Relative coordinates of corresponding point B on the connected image (NEW # XB,
NEW # YB) Note that by making these variable groups into an expandable array structure or list structure, it becomes possible to receive input of three or more corresponding points by the operator.

FIG. 9A is a diagram showing the corresponding points A and B input in this manner. The MPU 12 calculates, for each coordinate system of the two images, the gradients OLD # γ and NEW # γ of the vector AB connecting the corresponding points A and B as follows. OLD # γ = tan ^-1 [(OLD # YB-OLD # YA) / (OLD # XB-OLD # XA)] + π ... [Equation 4] NEW # γ = tan ^-1 [(NEW # YB- NEW # YA) / (NEW # XB-NEW # XA)] + π (Equation 5) The MPU 12 calculates the relative angle (OLD # γ-NEW # γ) of the vector AB from the two inclinations thus obtained. ) Is calculated. M
The PU 12 obtains the rotation amount NEW # θ of the connected image by adding the rotation amount OLD # θ of the base image to the relative angle by an offset (S3 in FIG. 7).

NEW # θ = OLD # γ−NEW # γ + OLD # θ (Equation 6) When the base image is the first image selected in the connection work, the coordinate space of the base image is common. Since the coordinate space is used, the rotation amount OLD # θ in the above equation is set to zero. At this point, as shown in FIG. 9B, the arrangement of the base image in the common coordinate space has already been determined. Then, M
The PU 12 performs coordinate conversion for mapping the base image in the common coordinate space by using relative coordinates (OLD # X) of the corresponding point A input on the base image.
A, OLD # YA). As a result, the absolute coordinates (OLD # XA0, OLD # YA
0) is determined (S4 in FIG. 7).

OLD # XA0 = OLD # XA · cos (OLD # θ) −OLD # YA · sin (OLD # θ) + OLD # X0 [Expression 7] OLD # YA0 = OLD # XA · sin (OLD # θ) + OLD # YA · cos (OLD # θ) + OLD # Y0 (Equation 8) In the above equation, (OLD # X0, OLD # Y0) is the upper left of the base image in the common coordinate space. Point coordinates. Then, MPU12
Is the translation amount P (OLD # XA0-NEW # XA, OLD # YA0) of the corresponding point A from the distance between "the corresponding point A on the base image in the common coordinate space" and "the corresponding point A on the connected image". -NEW # YA) (S5 in FIG. 7).

Using the parallel movement amount P and the rotation amount NEW # θ of the corresponding point A obtained as described above, it becomes possible to perform coordinate transformation of the connected image on the common coordinate space. However,
The connected image on the memory 17 is stored with the corresponding point A as the origin and the upper left point as the origin. Therefore, in order to further increase the calculation efficiency when performing coordinate transformation of the connected image,
The “parallel movement amount P of the corresponding point A” is converted into “the upper left point parallel movement amount Q” in the following procedure.

First, the MPU 12 translates the upper left point of the connected image using the translation amount P, and then rotates the rotation point NEW # θ about the corresponding point A on the common coordinate space. As a result, as shown in the following expression, the absolute coordinates (NEW # X0, NEW # Y0) of the upper left point of the connected image in the common coordinate space are determined. NEW # X0 = -NEW # XA · cos (NEW # θ) + NEW # YA · sin (NEW # θ) + OLD # XA0 ・ ・ ・ [Equation 9] NEW # Y0 = -NEW # XA · sin (NEW # θ) -NEW # YA · cos (NEW # θ) + OLD # YA0 (Equation 10) FIG. 9C shows the absolute coordinates (NEW # X0, NEW # Y0) of the upper left point determined in this way. FIG.

By the above calculation, the translation amount Q of the upper left point
Is determined to be (NEW # X0, NEW # Y0) (S6 in FIG. 7). Here, the MPU 12 stores the “parallel movement amount Q of the upper left point” and the “rotation amount NEW # θ” obtained as described above in association with the connected image (S7 in FIG. 7). Subsequently, the MPU 12
A plurality of line segments passing through two or more corresponding points are determined, and the total area of unnecessary portions to be cut is calculated for each of these line segments. From the result of such calculation, the MPU 12 selects a line segment having the smallest total area of the unnecessary portion as the cut line (S8 in FIG. 7).

FIG. 10 is a diagram specifically explaining such selection of the cut line. FIG. 10 (a) to (c)
3 shows unnecessary portions (regions filled with white circles in the figure) cut by the line segments AB, BC, and CA, respectively. in this case,
The case where the unnecessary area is cut by the line segment AB in FIG. 10A has the smallest unnecessary area. Therefore, the line segment AB is selected as the cut line.

Unnecessary parts of the base image and the connected image are cut (erased from the memory 17) with the cut line selected as a boundary (S9 in FIG. 7). Here, MPU1
2 displays a dialog on the monitor screen, informs the operator of the progress of the connection work, and waits for an instruction from the operator.
If there is another image to be connected (YES in S10 in FIG. 7), the MPU 12 returns the operation to step S1 and repeats the connection operation.

On the other hand, when there is no other image to be connected (NO in FIG. 7S10), the MPU 12 obtains the image display angle α from the operator (S11 in FIG. 7). The MPU 12
The translation amount Q and the rotation amount NEW # θ of each image are converted by the image display angle α as shown in FIG. As a result, as shown in the following equation, the translation amount Q '(NEW # X0', NEW # Y0 ') and the rotation amount NEW # θ' of each image are determined in consideration of the image display angle α (S12 in FIG. 7). ).

NEW # X0 ′ = NEW # X0 · cos (α) −NEW # Y0 · sin (α) [Expression 11] NEW # Y0 ′ = NEW # X0 · sin (α) + NEW # Y0 · cos (α) [Expression 12] NEW # θ ′ = NEW # θ + α [Expression 13] Here, the MPU 12 calculates the coordinate change parameter [NEW # θ ′, Q ′] of each image. To perform the following coordinate transformation on each image.

(1) First, each image is rotationally transformed by an amount of rotation NEW # θ 'about the upper left point. (2) Each image which has been rotationally transformed is translated by a parallel movement amount Q'. By the conversion, the images are rearranged on the same address space (common coordinate space) of the memory 17 and become connected images (S13 in FIG. 7).

In the course of such coordinate conversion, if the pixel positions before and after the conversion do not correspond one-to-one, the resampling of the pixels is appropriately executed. MPU12
Stores the connected image thus created in the memory 17.
The data is transferred from above to the hard disk 16 and stored in the object-oriented database (S14 in FIG. 7). At this time,
The management table of the connected images in the object-oriented database contains the names of the individual image files used in the current connection processing and connection information of each image file (for example, the four corner coordinates of each image after the connection processing). Is stored.

Here, the MPU 12 displays a dialog on the monitor screen, informs the operator that the connection processing has been completed, and waits for a new instruction from the operator. If connection processing is performed for another image group (FIG. 7S
15 (YES side of 15), the MPU 12 returns the operation to step S1 after initializing the coordinate conversion parameters and the like (FIG. 7).
S16).

On the other hand, if the connection processing has been completed for the time being (NO in S15 in FIG. 7), the MPU 12 shifts to the display operation of the connected image in the following procedure. First, MPU
12 reads a partial range of the connected image from the object-oriented database of the hard disk 16 and transfers it to the image processing board 18. The image processing board 18 stores this transfer in an internal image buffer and initially displays it on the monitor screen (S17 in FIG. 8).

Here, the MPU 12 enters a standby state to wait for an input message from the operator. Where MP
The input messages interpreted and executed by U12 are, for example, "scroll instruction", "mouse click for object operation", "window end instruction", and the like. Hereinafter, these will be described in order. (Operation in Response to Scroll Instruction) If a scroll instruction message is input during the standby state (YE in FIG. 8S18)
(S side), the MPU 12 determines whether or not the connected image on the monitor screen can be scrolled in the indicated direction (S1 in FIG. 8).
9).

If the vicinity of the end of the image in the designated direction is displayed on the monitor screen (NO in S19 in FIG. 8),
The MPU 12 determines that scrolling is not possible. In this case, the management table for the connected images in the object-oriented database is searched to determine whether there is another connected image sharing the currently displayed image portion (S2 in FIG. 8).
0a). If there is no other connected image, MPU
No. 12 gives up the scroll operation and returns to the standby state. On the other hand, when there is another connected image, the MPU 12 switches the display to another connected image and continues the scroll operation (S20 in FIG. 8).

If it is determined in step S19 that scrolling is possible, the MPU 12 gives a scroll command to the image processing board 18. The image processing board 18 shifts the display range on the image buffer in response to the scroll command to realize scrolling movement on the monitor screen (S21 in FIG. 8). At this time, when the drawing after the scroll movement is not completed on the image processing board 18, a message notifying “occurrence of an invalid rectangle” is transmitted from the image processing board 18 to the MPU 12. MP
U12 transmits the image data for the invalid rectangle to the image processing board 1
8 to eliminate the invalid rectangle.

The operation described above (step S18 in FIG. 8)
-S21) realizes scroll display over a plurality of connected images. (Operation in Response to Mouse Click) When a mouse click is executed on the display image during the standby state (Y in S22 in FIG. 8).
(ES side), the MPU 12 searches the registration table in the object-oriented database, and determines whether or not the object has already been registered at the click location of the displayed connected image (S23 in FIG. 8).

If no object is registered at the click location (NO in S23 of FIG. 8), the MPU 1
2 displays a dialog for new object registration. On this dialog, the operator can register files (movie files, still image files, audio files, executable files, character files, etc.) to be newly registered as objects.
Select The MPU 12 newly registers the selected file in the registration table in the object-oriented database in association with the clicked position of the displayed connected image (S24 in FIG. 8).

On the other hand, if the object has already been registered at the click location (YES side of S23 in FIG. 8), the MPU
12 determines whether the current operation mode of the image connection program is the object execution mode (S2 in FIG. 8).
5). If in the object execution mode (FIG. 8S
(YES side of 25), the MPU 12 generates a child thread for executing the registered object. This child thread executes the registered object in synchronization with the scroll operation (steps S18 to S21 in FIG. 8) of the parent thread (S26 in FIG. 8).

On the other hand, in the case of the object editing mode (NO in S25 in FIG. 8), the MPU 12 displays a dialog for editing the object and accepts an instruction to edit the object (eg, editing, updating, deleting, etc.) from the operator. (FIG. 8S27). Through the above operation (steps S22 to S27 in FIG. 8), the operation of the object associated with the connected image can be performed.

(Window end instruction) When the window end message is received during the standby state (YES side in S28 in FIG. 8), the MPU 12 sequentially deletes each window and releases resources related to the image connection program. Execute (S29 in FIG. 8). After finishing such post-processing,
The MPU 12 exits from the message waiting state and ends the image connection program.

(Effects of Second Embodiment, etc.) With the operations described above, the second embodiment can provide the same effects as those of the first embodiment. Furthermore, in the second embodiment, a line segment in which the total area of unnecessary portions to be cut is as small as possible is selected as a cut line. Therefore, it is possible to minimize the disappearance of the image information after the connection processing.

In the second embodiment, a rotation process corresponding to the image display angle is executed during the image connection process.
Therefore, it is not necessary to perform the image rotation processing during the scroll display, and the speed of the scroll display can be further increased. Further, in the second embodiment,
Since the registered object can be executed by mouse click, an image connection system excellent in presenting various information is realized.

In the second embodiment, the object can be updated or newly registered by clicking the mouse, so that an image connection system excellent in flexible object operation is realized. In particular, in the second embodiment, coordinate conversion parameters [NEW # θ ′, Q ′] that can execute coordinate conversion of a connected image in the shortest procedure are used. Therefore, it is possible to further increase the calculation efficiency when performing coordinate transformation of the connected image.

[0102]

According to the first aspect of the invention, by inputting corresponding points on the screen while collating a plurality of images, the positional relationship between the plurality of images is simply and clearly input. It becomes possible. Therefore, in order to input the positional relationship between images, it is possible to realize an excellent man-machine interface without troublesome work such as calculating and inputting minute numerical values such as coordinate positions and inclinations of images one by one. .

On the other hand, the moving amount calculating section only needs to pay attention to at least two corresponding points, and can calculate the parallel moving amount and the rotating amount of the image with a small amount of calculation. Also,
Since the image arrangement unit performs the coordinate transformation of the image using the translation amount and the rotation amount, it is possible to perform appropriate connection processing even on images having different rotation angles.

In particular, there is no particular need to perform complicated residual calculation and feature information extraction processing as compared with the conventional example. Therefore, it is possible to greatly reduce the processing time of the image connection (however, a conventional algorithm may be used together for correction of the connection processing or the like). Furthermore, even for images (a) to (d) of types that are not suitable for conventional template matching or structure matching, it is possible to perform image connection without much problem.

(A) An image in which the gradation change of the image is small or gradual. (B) An image in which the gradation change of the image is extremely large or periodic. (C) The overlapping area of the image is small. Image with insufficient information for matching (d) Low resolution digital image

(Claim 2) In the invention according to claim 2,
The amount of translation and the amount of rotation are calculated under mild conditions such as "at least one corresponding point overlaps" and "the inclination of a vector connecting one corresponding point and another corresponding point coincides". Therefore, even if the corresponding points include some errors, it is possible to execute the connecting process of the images flexibly and reliably.

(Claim 3) In the invention according to claim 3,
Exact connection processing can be performed on a plurality of images having different scale ratios. In particular, since the scale relation between the images is obtained by using the corresponding points acquired from the position input unit, the operator does not need to calculate the scale ratio one by one and input it separately.

(Claim 4) In the invention according to claim 4,
Since a line segment passing through the corresponding point is used for the cut line, complicated processing for properly determining the cut line is not required. Also, the operator can control the setting of the cut line to some extent depending on the position of the corresponding point to be given.

(Claim 5) In the invention according to claim 5,
Since a line segment that is as orthogonal as possible to the connecting direction of the images is selected as the cut line, it is possible to cut the connecting portions of a plurality of images in an appropriate direction.

(Claim 6) In the invention according to claim 6,
A line segment in which the total area of unnecessary portions to be cut is as small as possible is selected as a cut line. By setting such a cut line, it is possible to suppress the disappearance of the image information after the connection processing.

(Claim 7) In the invention according to claim 7,
Since the rotation processing of the image according to the image display angle is completed during the image connection, the rotation processing of the image during the scroll display can be omitted. Therefore, the scroll display speed can be further increased.

(Claim 8) In the invention according to claim 8,
Since an object registered in the object-oriented database is selected and executed in response to an external position instruction, an image connection system excellent in presenting various information is realized.

(Claim 9) In the invention according to claim 9,
Since the object is updated or newly registered in the object-oriented database in response to an external position instruction, an image connection system excellent in flexible information operation is realized.

(Claim 10) The use of the recording medium according to claim 10 enables the computer to execute claim 1 on a computer.
The image connection system according to any one of Items 1 to 6, can be realized.

(Claim 11) By using the recording medium according to claim 11, it is possible to execute the method according to claim 7 on a computer.
The image connection system according to any one of the above items 9 to 9, can be realized.

[Brief description of the drawings]

FIG. 1 is a diagram showing an overall configuration of an image connection system.

FIG. 2 is a flowchart illustrating the operation of the first embodiment.

FIG. 3 is a diagram showing a display example of a base image and a connected image.

FIG. 4 is a diagram for explaining image connection processing;

FIG. 5 is a diagram for explaining an unnecessary part cutting process.

FIG. 6 is a diagram illustrating a case where images are connected in the vertical direction.

FIG. 7 is a flowchart (1/1/2) for explaining the operation of the second embodiment;
2).

FIG. 8 is a flowchart for explaining the operation of the second embodiment (2 /
2).

FIG. 9 is a diagram for explaining image connection processing;

FIG. 10 is a diagram for explaining selection of a cut line.

FIG. 11 is a diagram illustrating coordinate conversion based on an image display angle α.

[Explanation of symbols]

 Reference Signs List 11 computer 12 MPU 13 input device 16 hard disk 17 memory 18 image processing board 19 monitor 20 interface board 21 image input device 22 CD-ROM drive device 23 CD-ROM

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] October 7, 1998 (1998.10.
7)

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0046

[Correction method] Change

[Correction contents]

The MPU 12 sequentially takes in these mouse click positions via the input device 13 and stores them in the following variables, respectively. (1) Relative coordinates of the corresponding point A on the base image (OLD_XA, OL
D_YA) (2) Relative coordinates (NEW_XA,
NEW_YA) (3) Relative coordinates of the corresponding point B on the base image (OLD_XB, OL
D_YB) (4) Relative coordinates (NEW_XB,
NEW_YB) By making these variable groups into an expandable array structure or list structure, it becomes possible to receive input of three or more corresponding points by the operator.

[Procedure amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0047

[Correction method] Change

[Correction contents]

FIG. 4A is a diagram showing the corresponding points A and B input as described above. The MPU 12 calculates, for each coordinate system of the two images, the gradients OLD_γ and NEW_γ of the vector AB connecting the corresponding points A and B as follows. OLD_γ = tan ^-1 [(OLD_YB-OLD_YA) / (OLD_XB-OLD_XA)] + π ... [Equation 1] NEW_γ = tan ^-1 [(NEW_YB-NEW_YA) / (NEW_XB-NEW_XA)] + π ... [Equation 2] The MPU 12 calculates the relative angle (OLD_γ-NEW_γ) of the vector AB from the two inclinations thus obtained, and sets the relative angle as the rotation amount NEW_θ of the connected image (S2 in FIG. 2).

[Procedure amendment 3]

[Document name to be amended] Statement

[Correction target item name] 0048

[Correction method] Change

[Correction contents]

Next, the MPU 12 determines the corresponding point A on the base image.
Then, the translation amount P (OLD_XA-NEW_XA, OLD_YA-NEW_YA) of the corresponding point A is calculated from the distance between the corresponding point A and the corresponding point A on the connected image (S3 in FIG. 2).

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0049

[Correction method] Change

[Correction contents]

Further, the MPU 12 calculates the scale ratio S from the ratio between the magnitudes of the vectors AB in the two images as follows (S4 in FIG. 2). S = √ [[(OLD_YB-OLD_YA) ² + (OLD_XB-OLD_XA) ² ] / [(NEW_YB-NEW_YA) ² + (NEW_XB-NEW_XA) ² ]]… [Equation 3] By the calculation, all parameters necessary for performing coordinate conversion of the connected image are prepared.

[Procedure amendment 5]

[Document name to be amended] Statement

[Correction target item name] 0053

[Correction method] Change

[Correction contents]

Subsequently, the MPU 12 sequentially executes coordinate conversion of the connected image. First, as shown in FIG. 4B, the MPU 12 rotates the connected image around the corresponding point A by the amount of rotation.
Rotate by NEW_θ (S8 in FIG. 2). Then, MPU1
2 reduces (enlarges) the connected image by the scale ratio S around the corresponding point A, as shown in FIG.
9).

[Procedure amendment 6]

[Document name to be amended] Statement

[Correction target item name] 0056

[Correction method] Change

[Correction contents]

On the other hand, on the image connection system side, by acquiring at least two corresponding points from the input device 13, the parallel movement amount P and the rotation amount NEW_θ of the image can be calculated. Therefore, it is possible to perform accurate connection processing even for images having different rotation angles. In particular,
In the process of such concatenation processing, there is no need to perform complicated residual calculation and feature information extraction processing as in the related art. Therefore, the processing time of the image connection can be significantly reduced as compared with the conventional example. Further, even for an image of a type for which conventional template matching or structure matching is not suitable, image connection can be performed without any problem.

[Procedure amendment 7]

[Document name to be amended] Statement

[Correction target item name] 0057

[Correction method] Change

[Correction contents]

Furthermore, in the first embodiment, the gradual “the one corresponding point A at least overlaps” and the “gradient of the vector connecting the one corresponding point A and the other corresponding point B match” are described. Under the conditions, the translation amount P and the rotation amount NEW_θ are calculated. Accordingly, even if the positions of the corresponding points A and B include some error, the image connection processing can be performed flexibly and reliably.

[Procedure amendment 8]

[Document name to be amended] Statement

[Correction target item name] 0062

[Correction method] Change

[Correction contents]

In the above-described first embodiment, since the connection direction of the images is processed as a rough direction such as up and down, left and right, and diagonal, the cut line selection processing can be greatly simplified. it can. In practice, such a simple process is often sufficient. However, it is not necessary to be limited to such processing. For example, the connection direction may be obtained in detail from the inclination of the parallel movement amount P. Also,
Based on the parallel movement amount P and the rotation amount NEW_θ, the “slope of the line connecting the center point of the base image and the center point of the connected image” is calculated in the common coordinate space, and the connection direction is determined in detail from the slope. You may ask.

[Procedure amendment 9]

[Document name to be amended] Statement

[Correction target item name] 0064

[Correction method] Change

[Correction contents]

In particular, the parameters (P, NEW_θ, S) used for the coordinate transformations (1) to (3) can be calculated from the two corresponding points in an almost shortest efficient calculation procedure. It is possible. Therefore, in the first embodiment, by performing the coordinate conversion procedure (1) to (3), it becomes possible to execute the parameter calculation for the coordinate conversion at a particularly high speed.

[Procedure amendment 10]

[Document name to be amended] Statement

[Correction target item name] 0076

[Correction method] Change

[Correction contents]

(1) Relative coordinates of corresponding point A on base image (OLD_XA, OLD_YA) (2) Relative coordinates of corresponding point A on connected image (NEW_XA,
NEW_YA) (3) Relative coordinates of the corresponding point B on the base image (OLD_XB, OL
D_YB) (4) Relative coordinates (NEW_XB,
NEW_YB) By making these variable groups into an expandable array structure or list structure, it becomes possible to receive input of three or more corresponding points by the operator.

[Procedure amendment 11]

[Document name to be amended] Statement

[Correction target item name] 0077

[Correction method] Change

[Correction contents]

FIG. 9A is a diagram showing the corresponding points A and B input in this manner. The MPU 12 calculates, for each coordinate system of the two images, the gradients OLD_γ and NEW_γ of the vector AB connecting the corresponding points A and B as follows. OLD_γ = tan ^-1 [(OLD_YB-OLD_YA) / (OLD_XB-OLD_XA)] + π ... [Equation 4] NEW_γ = tan ^-1 [(NEW_YB-NEW_YA) / (NEW_XB-NEW_XA)] + π ... [Equation 5] The MPU 12 calculates the relative angle (OLD_γ-NEW_γ) of the vector AB from the two inclinations thus obtained. M
The PU 12 obtains the rotation amount NEW_θ of the connected image by offset-adding the rotation amount OLD_θ of the base image to this relative angle (S3 in FIG. 7).

[Procedure amendment 12]

[Document name to be amended] Statement

[Correction target item name] 0078

[Correction method] Change

[Correction contents]

NEW_θ = OLD_γ−NEW_γ + OLD_θ (Equation 6) In the case where the base image is the first image selected in the connection work, the coordinate space of the base image is a common coordinate space. The middle rotation amount OLD_θ is set to zero. At this point, as shown in FIG. 9B, the arrangement of the base image in the common coordinate space has already been determined. Then, M
The PU 12 performs coordinate conversion for mapping the base image in the common coordinate space by using the relative coordinates (OLD_X) of the corresponding point A input on the base image.
A, OLD_YA). As a result, as shown in the following equation, the absolute coordinates (OLD_XA0, OLD_YA
0) is determined (S4 in FIG. 7).

[Procedure amendment 13]

[Document name to be amended] Statement

[Correction target item name] 0079

[Correction method] Change

[Correction contents]

OLD_XA0 = OLD_XA · cos (OLD_θ) −OLD_YA · sin (OLD_θ) + OLD_X0 ・ ・ ・ [Equation 7] OLD_YA0 = OLD_XA · sin (OLD_θ) + OLD_YA · cos (OLD_θ) + OLD_Y0 ・ ・ ・ [Equation 8] Note that (OLD_X0, OLD_Y0) in the above expression is the upper left point coordinate of the base image in the common coordinate space. Then, MPU12
Calculates the translation amount P (OLD_XA0-NEW_XA, OLD_YA0-NEW_YA) of the corresponding point A from the distance between the “corresponding point A on the base image in the common coordinate space” and the “corresponding point A on the connected image” ( (FIG. 7S5).

[Procedure amendment 14]

[Document name to be amended] Statement

[Correction target item name] 0080

[Correction method] Change

[Correction contents]

Using the parallel movement amount P and the rotation amount NEW_θ of the corresponding point A obtained as described above, it is possible to transform the connected image into a common coordinate space. However,
The connected image on the memory 17 is stored with the corresponding point A as the origin and the upper left point as the origin. Therefore, in order to further increase the calculation efficiency when performing coordinate transformation of the connected image,
The “parallel movement amount P of the corresponding point A” is converted into “the upper left point parallel movement amount Q” in the following procedure.

[Procedure amendment 15]

[Document name to be amended] Statement

[Correction target item name] 0081

[Correction method] Change

[Correction contents]

First, the MPU 12 translates the upper left point of the connected image using the translation amount P, and then rotates the rotation point NEW_θ about the corresponding point A on the common coordinate space. As a result, as shown in the following equation, the absolute coordinates (NEW_X0, NEW_Y0) of the upper left point of the connected image in the common coordinate space are determined. NEW_X0 = -NEW_XA · cos (NEW_θ) + NEW_YA · sin (NEW_θ) + OLD_XA0… [Equation 9] NEW_Y0 = -NEW_XA · sin (NEW_θ) -NEW_YA · cos (NEW_θ) + OLD_YA0… [Equation 10] FIG. 9C is a diagram showing the absolute coordinates (NEW_X0, NEW_Y0) of the upper left point determined in this way.

[Procedure amendment 16]

[Document name to be amended] Statement

[Correction target item name]

[Correction method] Change

[Correction contents]

By the above calculation, the translation amount Q of the upper left point
Is determined to be (NEW_X0, NEW_Y0) (S6 in FIG. 7). Here, the MPU 12 stores the “parallel movement amount Q of the upper left point” and the “rotation amount NEW_θ” obtained as described above in association with the connected image (S7 in FIG. 7). Subsequently, the MPU 12
A plurality of line segments passing through two or more corresponding points are determined, and the total area of unnecessary portions to be cut is calculated for each of these line segments. From the result of such calculation, the MPU 12 selects a line segment having the smallest total area of the unnecessary portion as the cut line (S8 in FIG. 7).

[Procedure amendment 17]

[Document name to be amended] Statement

[Correction target item name] 0085

[Correction method] Change

[Correction contents]

On the other hand, when there is no other image to be connected (NO in FIG. 7S10), the MPU 12 obtains the image display angle α from the operator (S11 in FIG. 7). The MPU 12
The translation amount Q and the rotation amount NEW_θ of each image are converted by the image display angle α as shown in FIG. As a result, as shown in the following equation, the translation amount Q ′ (NEW_X0 ′, NEW_Y0 ′) and the rotation amount NEW_θ ′ of each image are determined in consideration of the image display angle α (S12 in FIG. 7).

[Procedure amendment 18]

[Document name to be amended] Statement

[Correction target item name] 008

[Correction method] Change

[Correction contents]

NEW_X0 ′ = NEW_X0 · cos (α) −NEW_Y0 · sin (α) [Expression 11] NEW_Y0 ′ = NEW_X0 · sin (α) + NEW_Y0 · cos (α) [Expression 12] NEW_θ '= NEW_θ + α (Equation 13) Here, the MPU 12 performs the following coordinate conversion on each image using the coordinate change parameter [NEW_θ', Q '] of each image.

[Procedure amendment 19]

[Document name to be amended] Statement

[Correction target item name] 0087

[Correction method] Change

[Correction contents]

(1) First, each image is rotationally transformed by an amount of rotation NEW_θ ′ about the upper left point. (2) Each image that has been rotationally transformed is translated by a parallel movement amount Q ′. Each image is rearranged on the same address space (common coordinate space) of the memory 17 and becomes a connected image (S13 in FIG. 7).

[Procedure amendment 20]

[Document name to be amended] Statement

[Correction target item name] 0101

[Correction method] Change

[Correction contents]

In the second embodiment, the object can be updated or newly registered by clicking the mouse, so that an image connection system excellent in flexible object operation is realized. In particular, in the second embodiment, coordinate conversion parameters [NEW_θ ′, Q ′] that can execute coordinate conversion of a connected image in the shortest procedure are used. Therefore, it is possible to further increase the calculation efficiency when performing coordinate transformation of the connected image. ────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission Date] August 19, 1999 (1999.8.1)
9)

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Claims

[Correction method] Change

[Correction contents]

[Claims]

[Procedure amendment 2]

[Document name to be amended] Statement

[Correction target item name] Detailed description of the invention

[Correction method] Change

[Correction contents]

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image connection system for performing connection processing of a plurality of images, and a recording medium storing an image connection program. In particular, the present invention relates to a technology of a man-machine interface for easily and reliably inputting a positional relationship between these images.

[0002]

2. Description of the Related Art Conventionally, an algorithm for performing template matching or structure matching has been known as an algorithm of an image connection system. Among these, the image connection system using template matching executes the image connection processing in the following procedure.

(Step 1) Image data in a partial range is cut out from an image to be connected (hereinafter, referred to as a "connected image") to create a template image. (Step 2) A partial range is initially set within a range of a reference image (hereinafter, referred to as a “base image”). (Step 3) A difference for each pixel is calculated between a partial range of the base image and the template image, and a sum of absolute values of these pixel differences (so-called “residual”) is calculated. If the intermediate result of the residual exceeds a predetermined threshold during this calculation,
It is determined that there is no possibility that the partial range of the base image matches the template image, and the calculation of the residual is terminated at that time.

(Step 4) It is determined whether or not the final result of the residual is equal to or smaller than an allowable value. If it is larger than the allowable value, it is determined that the partial range of the base image does not match the template image. In this case, the operation returns to step 3 after the setting of the partial range is slightly shifted on the base image. On the other hand, if it is equal to or smaller than the allowable value, it is determined that the partial range of the base image matches the template image, and the operation shifts to step S5.

(Step 5) The connected image is translated so that the template image overlaps a part of the base image,
The base image and the connected image are connected. In addition to the algorithm for calculating the residual as described above (residual sequential test method), an algorithm for judging image coincidence from a cross-correlation coefficient between a base image and a template image is known (for details, see 19).
The University of Tokyo Press "Image Analysis Handbook" published in 1991
707-712).

On the other hand, as an image connection system using structured matching, there is known an image connection system that extracts characteristic information such as a branch point from an image and performs image position adjustment based on the characteristic information (detailed description). Is described in the above-mentioned “Image Analysis Handbook”, pp. 713 to 746). Also, recently, a method using FFT transform or Gabor Wavelet transform for extracting feature information is known.

[0007]

In the conventional example using the template matching described above, if there is a difference in the rotation angle or a difference in the scale ratio between the base image and the connected image, the matching judgment is made. In doing so, the rotation angle and scale must be changed many times. For this reason, there has been a problem that the number of calculations of the residual and the like is greatly increased.

On the other hand, in the conventional example using the structured matching described above, an amount invariant to the rotation angle or the like (for example, an FFT transform coefficient calculated on a circle) can be used as the feature information. Therefore, in theory, such an alignment of the feature information makes it possible to obtain the amount of translation of the image without being affected by the difference in the rotation angle.

However, in a digital image, if the rotation angle and the image size are different, the pixel positions between the images do not correspond one-to-one. Therefore, for a low-resolution digital image, there is a problem that the invariance of feature information is not so established, and it is difficult to apply structured matching.

In addition, the following (a) to (a) are common to the conventional examples of the template matching and the structured matching described above.
There is also a problem that the image as shown in FIG. (A) The case where the gradation change of the image is small or gradual (b) The case where the gradation change of the image is extremely large or periodic (c) The overlapping area of the image is small, and Therefore, in order to solve the above-described problems, a man-machine interface for simply and surely inputting a positional relationship between images is provided in the inventions according to claims 1 to 3. It is an object of the present invention to provide an image linking system provided with the same.

[0011] In the invention described in claim 1, the unnecessary portion of each image, and an object thereof is to provide an image articulated system which can be cut to the proper orientation. In the invention described in claim 2, and an object thereof is to provide an image articulated system capable of reducing the disappearance of the image information after concatenation process.

[0012] In the invention of claim 3 is intended to provide in performing scroll display while changing the display angle of the image, the image connecting system capable of speeding up the speed of the scroll display. In the invention described in claim 4 and 5, it was added to object-oriented database, and an object thereof is to provide an image articulation system specific forms.

According to the sixth aspect of the present invention, the first and second aspects are provided.
It is an object of the present invention to provide a recording medium recording an image connection program for realizing the image connection system described in 1 above on a computer. In the invention described in claim 7, wherein
It is an object of the present invention to provide a recording medium storing an image connection program for realizing the image connection system according to any one of Items 3 to 5 on a computer.

[0014]

Means for Solving the Problems ( Common to Claims 1 to 7)
That structure) common structure of claims 1 to 7, in the image connecting system for connecting a plurality of images, receiving a position input on the screen a plurality of images together or sequentially displayed, among a plurality of images A position input unit that captures at least two overlapping corresponding points, and a translation amount and a rotation amount for overlapping the “corresponding point of another image” with the “corresponding point of one image” obtained by the position input unit. A movement amount calculation unit that calculates
An image arranging unit is provided that performs coordinate conversion of other images using the parallel movement amount and the rotation amount calculated by the movement amount calculation unit, and arranges a plurality of images on a common coordinate space.

In such a configuration, a position input section is newly provided as a man-machine interface. as a result,
The operator can find the corresponding points overlapping between the images while collating the plurality of images on the screen, and can perform an input operation easily and quickly. On the other hand, the movement amount calculation unit calculates the amount of translation and rotation of the image from at least two corresponding points acquired from the position input unit. The image arranging unit executes the coordinate transformation of the image using the translation amount and the rotation amount. Therefore, appropriate connection processing can be performed on images having different rotation angles.

( Composition common to Claims 1 and 2 )
The common configuration of 1 and 2 is that, in the above configuration, when arranging each image on the common coordinate space , the image arrangement section cuts an unnecessary portion of each image with a line segment passing through the corresponding point obtained by the position input section. It is characterized by the following.

Incidentally, regarding the processing of the overlapping area caused by the image concatenation, the processing of calculating the average value or the intermediate value (the case where three or more images overlap) between the overlapping pixels and the processing of the image (mainly the overlapping area) are performed. A process of dividing and cutting one as an unnecessary part can be considered. In particular, in the case of the latter processing, it is necessary to appropriately set a cut line that reliably divides the overlapping area, and usually, complicated calculations and determination processing are indispensable.

However, in the common configuration of claims 1 and 2, a line segment passing through a corresponding point is simply used as a cut line. Generally, the corresponding points are guaranteed to be located in the overlapping area between the images. Therefore, the line segment passing through the corresponding point always divides the overlapping area, and is a setting that is approximately appropriate as a cut line. Therefore, it is possible to set a cut line very easily and cut unnecessary portions of each image without much complicated calculation and determination processing.

Further, the operator can control the setting of the cut line to some extent by the position of the corresponding point to be given. Therefore, it is possible to realize an excellent man-machine interface that can reflect the intention of the operator on the setting of the cut line.

The "line segment" described in the common configuration of the first and second aspects is not limited to a line segment passing through only two corresponding points. For example, a line segment having an appropriate inclination passing through one corresponding point or a line segment passing through three or more corresponding points may be used. The "line segment" described in claims 1 and 2
Is not limited to a straight line. It may be a bent line or a curve (for example, a Bezier curve, a multi-degree curve, a hyperbola, etc.).

( Claim 1 ) The invention according to claim 1 is characterized in that
In the common configuration of claims 1 and 2 described above, a line segment whose angle with the connecting direction of the image is close to a right angle is selected from a plurality of line segments passing through two or more corresponding points, and each line segment is selected using the selected line segment. It is characterized in that unnecessary portions of the image are cut.

In the above configuration, a line segment that is as orthogonal as possible to the direction in which the images are connected is selected and set as a cut line. By setting such a cut line, an image (mainly, an overlapping area) is divided in a connecting direction, and one of the divided areas is cut as an unnecessary part. As a result, the connected portions of the plurality of images are cut in an appropriate direction. ( Claim 2 ) The invention described in claim 2 is the above-described claim.
In the common configuration of 1 and 2, a line segment having the smallest total area of the unnecessary portion to be cut is selected from a plurality of line segments passing through two or more corresponding points, and the unnecessary portion of each image is selected using the selected line segment. Is characterized by cutting.

In general, when only an overlapping area is divided by a cut line and one of them is cut as an unnecessary part, the image information after the concatenation processing does not disappear at all. However, cut the entire image at the cut line,
If one of them is cut as an unnecessary part, image information other than the overlapping area may be cut as an unnecessary part.

Therefore, in the above configuration, a line segment in which the total area of unnecessary portions to be cut is as small as possible is selected as a cut line. By setting such a cut line,
It is possible to suppress the disappearance of the image information after the connection processing. ( Claim 3 ) The invention described in claim 3 is the above-described claim.
In the common configuration of 1 to 7, a plurality of images arranged in the common coordinate space are rotated according to an externally input image display angle, and are stored as a connected image. A scroll display unit for sequentially displaying the stored connected images in accordance with the scroll direction and executing scroll display over a plurality of connected images.

In such a configuration, since the image rotation processing according to the image display angle is completed during the image connection, the image rotation processing during the scroll display can be omitted. Therefore, the scroll display speed can be further increased. ( Claim 4 ) The invention according to claim 4 is the image linking system according to claim 3 , wherein the image storage section associates the linked image being stored with moving image information, still image information, audio information, An object-oriented database in which objects containing at least one of command information and character information can be registered, and an object is selected from the object-oriented database in response to an externally provided position indication on a screen, and the selected object is displayed as an image. And an object execution unit that executes the program at the same timing as described above.

( Claim 5 ) The invention described in claim 5 is a contract
In the image connection system according to claim 3 , the image storage unit registers an object including at least one of moving image information, still image information, audio information, command information, and character information in association with the stored connected image. A possible object-oriented database and an object updating unit that updates the object-oriented database by executing at least one of the following (1) and (2) in response to an externally given position indication on the screen. Features.

(1) A process of selecting a registered object corresponding to the position indication from the object-oriented database, and executing one of editing, updating, and deleting for the selected registered object. (2) The position A process for newly registering an object in the object-oriented database in association with an instruction. ( Claim 6 ) A recording medium according to claim 6 , wherein the recording medium is a computer, and the recording medium according to any one of claims 1 to 2 . An image connection program for functioning as a position input unit, a movement amount calculation unit, and an image arrangement unit is recorded.

By executing such an image connection program on a computer, the image connection system according to any one of claims 1 to 2 can be realized. Recording medium according to (Claim 7) according to claim 7, the computer, position input unit according to any one of claims 3 to claim 5, the movement amount calculation unit, the image placement block, the image storage unit,
An image connection program for functioning as a scroll display unit is recorded.

By executing such an image connection program on a computer, an image connection system according to any one of claims 3 to 5 can be realized.

[0030]

Embodiments of the present invention will be described below with reference to the drawings.

[0031] <First Embodiment> In the first embodiment, 請
This is an embodiment corresponding to the inventions described in claims 1 and 6 .

FIG. 1 is a diagram showing the overall configuration of an image connection system using a computer 11. In FIG. 1, an MPU (microprocessor) 12 is provided inside a computer 11. An input device 13 including a keyboard and a mouse, a hard disk 16, a memory 17, an image processing board 18, and an interface board 20 are connected to the MPU 12. A monitor 19 is connected to an image output terminal of the image processing board 18. On the other hand, an image input device 21 such as a scanner or an electronic camera is connected to the interface board 20.

On the other hand, a CD-ROM drive device 22 is connected to the MPU 12. Into the CD-ROM drive 22, a CD-ROM 23 in which an image connection program and its installation program are recorded is inserted. With the installation program in the CD-ROM 23, the MPU 12 develops the image connection program in the CD-ROM 23 and stores it in the hard disk 16 in an executable state.

(Correspondence between the present invention and the first embodiment)
Hereinafter, the correspondence relationship with the configuration requirements of the present invention will be described. First, regarding the correspondence between the common configuration of the first to seventh embodiments and the first embodiment, the position input unit includes the input device 13 and the M.
The movement amount calculation unit corresponds to “the function of displaying the base image and the connected image and acquiring the mouse input of the corresponding point” of the PU 12, and the MPU 12 “the rotation amount of the connected image based on the two or more corresponding points”. And the function to calculate the amount of translation).
The image arranging unit corresponds to the “function of converting the coordinate of the connected image according to the rotation amount and the parallel movement amount and arranging the base image and the connected image on the same address space of the memory 17” of the MPU 12.

Further, regarding the correspondence between the invention described in claim 1 and the first embodiment, the image arranging section is configured by MPU 1
2 corresponds to “a function of removing unnecessary portions of each image with a line segment that is as orthogonal as possible to the connecting direction among the line segments passing through the corresponding points”.

Further, as for the correspondence between the invention described in claim 6 and the first embodiment, the recording medium is a CD-ROM.
Corresponding to 23. (Operation of First Embodiment) The operation of the first embodiment will be described below. FIG. 2 is a flowchart illustrating the operation of the first embodiment.

First, when the image connection program is started in the computer 11, the MPU 12 creates a dialog on the monitor screen. FIG. 3 is a diagram showing a display example of such a dialog. In FIG. 3, two windows are arranged in a dialog. In these windows, the base image and the connected image input from the image input device 21 and the like are displayed, respectively. With such a dialog displayed, the MPU 12 waits for a mouse input by the operator (S1 in FIG. 2).

The operator compares two images on the dialog and determines at least two overlapping corresponding points between the two images. For example, the two images displayed in FIG. 3 are map images. The operator finds a crossroad common to these map images, and determines the upper right corner and the lower right corner of the intersection of the crossroads as the corresponding points A and B, respectively. The operator determines the corresponding points A and B of the two images determined in this way.
The mouse pointer over and then perform a mouse click.

The MPU 12 sequentially takes in these mouse click positions via the input device 13 and stores them in the following variables. (1) Relative coordinates of the corresponding point A on the base image (OLD_XA, OL
D_YA) (2) Relative coordinates (NEW_XA,
NEW_YA) (3) Relative coordinates of the corresponding point B on the base image (OLD_XB, OL
D_YB) (4) Relative coordinates (NEW_XB,
NEW_YB) By making these variable groups into an expandable array structure or list structure, it becomes possible to receive input of three or more corresponding points by the operator.

FIG. 4A is a diagram showing the corresponding points A and B input in this manner. The MPU 12 calculates, for each coordinate system of the two images, the gradients OLD_γ and NEW_γ of the vector AB connecting the corresponding points A and B as follows. OLD_γ = tan ^-1 [(OLD_YB-OLD_YA) / (OLD_XB-OLD_XA)] + π ... [Equation 1] NEW_γ = tan ^-1 [(NEW_YB-NEW_YA) / (NEW_XB-NEW_XA)] + π ... [Equation 2] The MPU 12 calculates the relative angle (OLD_γ-NEW_γ) of the vector AB from the two inclinations thus obtained, and sets the relative angle as the rotation amount NEW_θ of the connected image (S2 in FIG. 2).

Next, the MPU 12 determines the corresponding point A on the base image.
Then, the translation amount P (OLD_XA-NEW_XA, OLD_YA-NEW_YA) of the corresponding point A is calculated from the distance between the corresponding point A and the corresponding point A on the connected image (S3 in FIG. 2).

Further, the MPU 12 calculates the scale ratio S from the ratio of the magnitudes of the vectors AB in the two images as follows (S4 in FIG. 2). S = √ [[(OLD_YB-OLD_YA) ² + (OLD_XB-OLD_XA) ² ] / [(NEW_YB-NEW_YA) ² + (NEW_XB-NEW_XA) ² ]]… [Equation 3] By the calculation, all parameters necessary for performing coordinate conversion of the connected image are prepared.

Here, the MPU 12 determines whether there are three or more corresponding points (S5 in FIG. 2). If there are only two corresponding points, the MPU 12 shifts the operation to step S7 in order to set a cut line from these two corresponding points. On the other hand, when there are three or more corresponding points, some line segments passing through these corresponding points can be considered. Therefore, MPU12
Determines a plurality of line segments passing through two or more corresponding points, and selects a line segment that is as orthogonal as possible to the connecting direction of images as a cut line (S6 in FIG. 2).

Using the cut line thus selected, the MPU 12 divides the overlapping area between the images, and removes one of the overlapping areas from the memory 17 as an unnecessary part (FIG. 2S
7). FIG. 5 is a diagram for specifically explaining such an unnecessary portion cutting process. In FIG. 5A, three corresponding points A to C are input from the operator side. First, the MPU 12 selects 2 out of these corresponding points A to C.
Three types of line segments AB, BC, and CA connecting the corresponding points of the points are determined. Next, the MPU 12 selects, from among these line segments,
A line segment AB orthogonal to the image connecting direction (here, the left and right direction) is selected as much as possible and set as a cut line.

As shown in FIG. 5B, the MPU 12
Using this cut line, the overlapping area of each image is divided, and the different ones are stored in the memory 17 as unnecessary parts.
Remove from above. By removing such unnecessary portions, as shown in FIG. 5C, the connected portions of the image are appropriately processed without any loss of the connected image information.

Subsequently, the MPU 12 sequentially executes coordinate conversion of the connected image. First, as shown in FIG. 4B, the MPU 12 rotates the connected image around the corresponding point A by the amount of rotation.
Rotate by NEW_θ (S8 in FIG. 2). Then, MPU1
2 reduces (enlarges) the connected image by the scale ratio S around the corresponding point A, as shown in FIG.
9).

Further, as shown in FIG. 4D, the MPU 12 translates the connected image by the parallel movement amount P, and arranges the base image and the connected image in the same address space of the memory 17 (see FIG. 4D). 2S10). Note that these steps S8
If the pixel positions before and after the conversion do not correspond one-to-one in the coordinate conversion of to 画素, the pixel is resampled. Further, in order to reduce the deterioration of the connected image due to such resampling of pixels, the size of the connected image may be enlarged prior to the coordinate transformation.

(Effects of First Embodiment, etc.) According to the above-described series of processes, in the first embodiment, two images can be connected in a correct positional relationship. In particular, since the operator can check a plurality of images on the monitor screen, the operator can easily and quickly find the corresponding points overlapping between the images and immediately input them via the input device 13 such as a mouse. Can be.

On the other hand, on the image connection system side, by acquiring at least two corresponding points from the input device 13, the parallel movement amount P and the rotation amount NEW_θ of the image can be calculated. Therefore, it is possible to perform accurate connection processing even for images having different rotation angles. In particular,
In the process of such concatenation processing, there is no need to perform complicated residual calculation and feature information extraction processing as in the related art. Therefore, the processing time of the image connection can be significantly reduced as compared with the conventional example. Further, even for an image of a type for which conventional template matching or structure matching is not suitable, image connection can be performed without any problem.

Furthermore, in the first embodiment, the gradual “the one corresponding point A at least overlaps” and the “gradient of the vector connecting the one corresponding point A and the other corresponding point B match” are described. Under the conditions, the translation amount P and the rotation amount NEW_θ are calculated. Accordingly, even if the positions of the corresponding points A and B include some error, the image connection processing can be performed flexibly and reliably.

In the first embodiment, the scale ratio S between images is calculated based on the ratio of the intervals between the corresponding points AB.
Therefore, it is possible to perform an accurate connection process on a plurality of images having different scale ratios. In particular, since the scale ratio between the images is obtained by diverting the input information of the corresponding points, there is no need for the operator to calculate the scale ratio one by one and input it separately.

Further, in the first embodiment, since a line segment passing through the corresponding point is used as the cut line, complicated processing for properly determining the cut line is not required. Also,
On the operator side, the setting of the cut line can be controlled to some extent by the position of the corresponding point to be given. In particular, in the first embodiment, since a line segment that is as orthogonal as possible to the connecting direction of the images is selected as the cut line, it is possible to cut the connecting portions of a plurality of images in an appropriate direction.

In the first embodiment described above, a case where two images are connected is described, but the present invention is not limited to this. For example, at least two overlapping corresponding points may be input for each of three or more images. In such a case, three or more images can be connected at a time. In the above-described first embodiment, the cut line is selected from the line segments passing through the two corresponding points.
It is not limited to this. For example, one or more corresponding points may be selected, and a line segment passing through the selected corresponding point may be used as the cut line.

Furthermore, in the first embodiment described above, the case where images are connected in the left-right direction has been described, but the present invention is not limited to this. For example, as shown in FIG. 6, images may be connected in the vertical direction. FIG.
In the case shown in (1), a line segment AB that is as orthogonal as possible to the direction in which the images are connected (in this case, the vertical direction) may be selected as the cut line.

In the above-described first embodiment, since the connecting direction of the images is processed as a rough direction such as up and down, left and right, and diagonal, the cut line selection processing can be greatly simplified. it can. In practice, such a simple process is often sufficient. However, it is not necessary to be limited to such processing. For example, the connection direction may be obtained in detail from the inclination of the parallel movement amount P. Also,
Based on the parallel movement amount P and the rotation amount NEW_θ, the “slope of the line connecting the center point of the base image and the center point of the connected image” is calculated in the common coordinate space, and the connection direction is determined in detail from the slope. You may ask.

Further, in the first embodiment described above, (1) rotation around the corresponding point A (2) enlargement / reduction around the corresponding point A (3) translation in the order of the distance of the corresponding point A The coordinate transformation of the connected image is being executed. However, it is a matter of course that a conversion formula for performing the coordinate conversions (1) to (3) at once may be created and the coordinate conversion may be completed at once.

In particular, the parameters (P, NEW_θ, S) used for the coordinate transformations (1) to (3) can be calculated from the two corresponding points in an almost shortest efficient calculation procedure. It is possible. Therefore, in the first embodiment, by performing the coordinate conversion procedure (1) to (3), it becomes possible to execute the parameter calculation for the coordinate conversion at a particularly high speed.

However, the procedure of coordinate conversion is as follows (1)
It is not limited only to (3). In general, you can change the order of coordinate transformation, shift the center of rotation transformation,
Equivalent coordinate transformation can be obtained indefinitely by shifting the center of enlargement / reduction transformation or adding invalid coordinate transformation. Therefore, it is of course possible to use any of these equivalent coordinate transformations.

In the first embodiment, the CD-ROM 23 is used as a recording medium for the image connection program. However, the type of the recording medium is not limited to this. In general, any recording medium may be used as long as it is a machine-readable medium. Further, in the above-described first embodiment, the base image and the connected image are simply arranged and displayed together. Such a display method has an advantage that both images can be easily collated, and a corresponding point can be found quickly and easily. However, the present invention is not limited to this display method. For example, the base image and the connected image may be displayed sequentially. In addition, the base image and the connected image may be partially enlarged and displayed according to an instruction from the operator. According to such a partially enlarged display,
It is possible to further improve the input accuracy of the corresponding points.

Next, another embodiment will be described. <Second Embodiment> A second embodiment is defined in claims 2 and 3.
This is an embodiment corresponding to the inventions described in 4 , 5, and 7 . Note that the system configuration of the second embodiment is the same as that of the first embodiment (FIG. 1), and thus redundant description is omitted here.

(Correspondence between the present invention and the second embodiment)
Hereinafter, the correspondence relationship with the configuration requirements of the present invention will be described. First, regarding the correspondence between the common configuration of the first to seventh aspects and the second embodiment, the position input unit includes the input device 13 and the M.
The movement amount calculation unit corresponds to “the function of displaying the base image and the connected image and acquiring the mouse input of the corresponding point” of the PU 12, and the MPU 12 “the rotation amount of the connected image based on the two or more corresponding points”. And the function to calculate the amount of translation).
The image arranging unit corresponds to the “function of converting the coordinate of the connected image according to the rotation amount and the parallel movement amount and arranging the base image and the connected image on the same address space of the memory 17” of the MPU 12.

[0062] Regarding the correspondence between the invention described in claim 2 and the second embodiment, the image arranging unit is an MPU 12
The “function of removing unnecessary portions of each image using a line segment having the smallest total area of unnecessary portions among the line segments passing through the corresponding points”. Further, the invention according to claim 3 and the second invention
The image storage unit corresponds to the hard disk 16 and the “function of storing the connected image rotated according to the image display angle in the hard disk 16” of the MPU 12, and the scroll display unit corresponds to the MPU 12.
(A function of executing scroll display while sequentially switching a plurality of connected images on the hard disk 16).

As for the correspondence between the invention described in claim 4 and the second embodiment, the object-oriented database corresponds to the “function of registering and managing connected images and objects” of the hard disk 16 and the MPU 12. The object execution unit corresponds to the “function of executing an object in response to a mouse input” of the MPU 12.

Further, regarding the correspondence between the invention described in claim 5 and the second embodiment, the object-oriented database corresponds to the “function for registering and managing connected images and objects” of the hard disk 16 and the MPU 12. The object update unit corresponds to the “function of updating an object in response to a mouse input” of the MPU 12. Regarding the correspondence between the invention described in claim 7 and the second embodiment, the recording medium corresponds to the CD-ROM 23.

(Operation of Second Embodiment) The operation of the second embodiment will be described below. FIGS. 7 and 8 are flowcharts for explaining the operation of the second embodiment. First, the operator starts an image connection program in the computer 11, and selects a base image and a connection image from a plurality of images captured from a scanner or the like (FIG. 7S
1).

Note that the base image here is an image selected first in the connection work, or an image whose connection in the common coordinate space is already known after the connection work has been completed. On the other hand, the connected image is an image that partially overlaps the base image. The MPU 12 displays the two images thus selected together on a monitor screen. In such a display state, the MPU 12 waits for a mouse input by the operator.

The operator compares two images on the monitor screen and determines at least two overlapping points between the two images. The operator places the mouse pointer on the corresponding point of each image determined in this way and executes a mouse click (S2 in FIG. 7). The MPU 12 sequentially takes in the positions of these mouse clicks via the input device 13 and stores them in the following variables.

(1) Relative coordinates of corresponding point A on base image (OLD_XA, OLD_YA) (2) Relative coordinates of corresponding point A on connected image (NEW_XA, OLD_XA, OLD_YA)
NEW_YA) (3) Relative coordinates of the corresponding point B on the base image (OLD_XB, OL
D_YB) (4) Relative coordinates (NEW_XB,
NEW_YB) By making these variable groups into an expandable array structure or list structure, it becomes possible to receive input of three or more corresponding points by the operator.

FIG. 9A is a diagram showing the corresponding points A and B input in this manner. The MPU 12 calculates, for each coordinate system of the two images, the gradients OLD_γ and NEW_γ of the vector AB connecting the corresponding points A and B as follows. OLD_γ = tan ^-1 [(OLD_YB-OLD_YA) / (OLD_XB-OLD_XA)] + π ... [Equation 4] NEW_γ = tan ^-1 [(NEW_YB-NEW_YA) / (NEW_XB-NEW_XA)] + π ... [Equation 5] The MPU 12 calculates the relative angle (OLD_γ-NEW_γ) of the vector AB from the two inclinations thus obtained. M
The PU 12 obtains the rotation amount NEW_θ of the connected image by offset-adding the rotation amount OLD_θ of the base image to this relative angle (S3 in FIG. 7).

NEW_θ = OLD_γ−NEW_γ + OLD_θ (Equation 6) In the case where the base image is the first image selected in the connection work, the coordinate space of the base image is the common coordinate space. The middle rotation amount OLD_θ is set to zero. At this point, as shown in FIG. 9B, the arrangement of the base image in the common coordinate space has already been determined. Then, M
The PU 12 performs coordinate conversion for mapping the base image in the common coordinate space by using the relative coordinates (OLD_X) of the corresponding point A input on the base image.
A, OLD_YA). As a result, as shown in the following equation, the absolute coordinates (OLD_XA0, OLD_YA
0) is determined (S4 in FIG. 7).

OLD_XA0 = OLD_XA · cos (OLD_θ) −OLD_YA · sin (OLD_θ) + OLD_X0 ・ ・ ・ [Equation 7] OLD_YA0 = OLD_XA · sin (OLD_θ) + OLD_YA · cos (OLD_θ) + OLD_Y0 ・ ・ ・ [Equation 8] Note that (OLD_X0, OLD_Y0) in the above expression is the upper left point coordinate of the base image in the common coordinate space. Then, MPU12
Calculates the translation amount P (OLD_XA0-NEW_XA, OLD_YA0-NEW_YA) of the corresponding point A from the distance between the “corresponding point A on the base image in the common coordinate space” and the “corresponding point A on the connected image” ( (FIG. 7S5).

Using the parallel movement amount P and the rotation amount NEW_θ of the corresponding point A obtained as described above, the connected image can be coordinate-transformed into the common coordinate space. However,
The connected image on the memory 17 is stored with the corresponding point A as the origin and the upper left point as the origin. Therefore, in order to further increase the calculation efficiency when performing coordinate transformation of the connected image,
The “parallel movement amount P of the corresponding point A” is converted into “the upper left point parallel movement amount Q” in the following procedure.

First, the MPU 12 translates the upper left point of the connected image using the translation amount P, and then rotates the rotation point NEW_θ about the corresponding point A on the common coordinate space. As a result, as shown in the following equation, the absolute coordinates (NEW_X0, NEW_Y0) of the upper left point of the connected image in the common coordinate space are determined. NEW_X0 = -NEW_XA · cos (NEW_θ) + NEW_YA · sin (NEW_θ) + OLD_XA0… [Equation 9] NEW_Y0 = -NEW_XA · sin (NEW_θ) -NEW_YA · cos (NEW_θ) + OLD_YA0… [Equation 10] FIG. 9C is a diagram showing the absolute coordinates (NEW_X0, NEW_Y0) of the upper left point determined in this way.

By the above calculation, the translation amount Q of the upper left point
Is determined to be (NEW_X0, NEW_Y0) (S6 in FIG. 7). Here, the MPU 12 stores the “parallel movement amount Q of the upper left point” and the “rotation amount NEW_θ” obtained as described above in association with the connected image (S7 in FIG. 7). Subsequently, the MPU 12
A plurality of line segments passing through two or more corresponding points are determined, and the total area of unnecessary portions to be cut is calculated for each of these line segments. From the result of such calculation, the MPU 12 selects a line segment having the smallest total area of the unnecessary portion as the cut line (S8 in FIG. 7).

FIG. 10 is a diagram specifically explaining such selection of a cut line. FIG. 10 (a) to (c)
3 shows unnecessary portions (regions filled with white circles in the figure) cut by the line segments AB, BC, and CA, respectively. in this case,
The case where the unnecessary area is cut by the line segment AB in FIG. 10A has the smallest unnecessary area. Therefore, the line segment AB is selected as the cut line.

Unnecessary portions of the base image and the connected image are cut (erased from the memory 17) with the cut line selected as a boundary (S9 in FIG. 7). Here, MPU1
2 displays a dialog on the monitor screen, informs the operator of the progress of the connection work, and waits for an instruction from the operator.
If there is another image to be connected (YES in S10 in FIG. 7), the MPU 12 returns the operation to step S1 and repeats the connection operation.

On the other hand, when there is no other image to be connected (NO in S10 in FIG. 7), the MPU 12 acquires the image display angle α from the operator (S11 in FIG. 7). The MPU 12
The translation amount Q and the rotation amount NEW_θ of each image are converted by the image display angle α as shown in FIG. As a result, as shown in the following equation, the translation amount Q ′ (NEW_X0 ′, NEW_Y0 ′) and the rotation amount NEW_θ ′ of each image are determined in consideration of the image display angle α (S12 in FIG. 7).

NEW_X0 ′ = NEW_X0 · cos (α) −NEW_Y0 · sin (α) [Expression 11] NEW_Y0 ′ = NEW_X0 · sin (α) + NEW_Y0 · cos (α) [Expression 12] NEW_θ '= NEW_θ + α (Equation 13) Here, the MPU 12 performs the following coordinate conversion on each image using the coordinate change parameter [NEW_θ', Q '] of each image.

(1) First, each image is rotationally transformed by an amount of rotation NEW_θ 'about the upper left point. (2) Each image which has been rotationally transformed is translated by a parallel movement amount Q'. Each image is rearranged on the same address space (common coordinate space) of the memory 17 and becomes a connected image (S13 in FIG. 7).

In the course of such coordinate conversion, if the pixel positions before and after the conversion do not correspond one-to-one, the resampling of the pixels is appropriately executed. MPU12
Stores the connected image thus created in the memory 17.
The data is transferred from above to the hard disk 16 and stored in the object-oriented database (S14 in FIG. 7). At this time,
The management table of the connected images in the object-oriented database contains the names of the individual image files used in the current connection processing and connection information of each image file (for example, the four corner coordinates of each image after the connection processing). Is stored.

Here, the MPU 12 displays a dialog on the monitor screen, informs the operator that the connection processing has been completed, and waits for a new instruction from the operator. If connection processing is performed for another image group (FIG. 7S
15 (YES side of 15), the MPU 12 returns the operation to step S1 after initializing the coordinate conversion parameters and the like (FIG. 7).
S16).

On the other hand, if the connection processing has been completed for the time being (NO in S15 in FIG. 7), the MPU 12 shifts to the display operation of the connected image in the following procedure. First, MPU
12 reads a partial range of the connected image from the object-oriented database of the hard disk 16 and transfers it to the image processing board 18. The image processing board 18 stores this transfer in an internal image buffer and initially displays it on the monitor screen (S17 in FIG. 8).

Here, the MPU 12 enters a standby state to wait for an input message from the operator. Where MP
The input messages interpreted and executed by U12 are, for example, "scroll instruction", "mouse click for object operation", "window end instruction", and the like. Hereinafter, these will be described in order. (Operation in Response to Scroll Instruction) If a scroll instruction message is input during the standby state (YE in FIG. 8S18)
(S side), the MPU 12 determines whether or not the connected image on the monitor screen can be scrolled in the indicated direction (S1 in FIG. 8).
9).

If the vicinity of the end of the image in the designated direction is displayed on the monitor screen (NO in S19 in FIG. 8),
The MPU 12 determines that scrolling is not possible. In this case, the management table for the connected images in the object-oriented database is searched to determine whether there is another connected image sharing the currently displayed image portion (S2 in FIG. 8).
0a). If there is no other connected image, MPU
No. 12 gives up the scroll operation and returns to the standby state. On the other hand, when there is another connected image, the MPU 12 switches the display to another connected image and continues the scroll operation (S20 in FIG. 8).

If it is determined in step S 19 that scrolling is possible, the MPU 12 gives a scroll command to the image processing board 18. The image processing board 18 shifts the display range on the image buffer in response to the scroll command to realize scrolling movement on the monitor screen (S21 in FIG. 8). At this time, when the drawing after the scroll movement is not completed on the image processing board 18, a message notifying “occurrence of an invalid rectangle” is transmitted from the image processing board 18 to the MPU 12. MP
U12 transmits the image data for the invalid rectangle to the image processing board 1
8 to eliminate the invalid rectangle.

The operation described above (step S18 in FIG. 8)
-S21) realizes scroll display over a plurality of connected images. (Operation in Response to Mouse Click) When a mouse click is executed on the display image during the standby state (Y in S22 in FIG. 8).
(ES side), the MPU 12 searches the registration table in the object-oriented database, and determines whether or not the object has already been registered at the click location of the displayed connected image (S23 in FIG. 8).

If no object is registered at the click location (NO in S23 of FIG. 8), the MPU 1
2 displays a dialog for new object registration. On this dialog, the operator can register files (movie files, still image files, audio files, executable files, character files, etc.) to be newly registered as objects.
Select The MPU 12 newly registers the selected file in the registration table in the object-oriented database in association with the clicked position of the displayed connected image (S24 in FIG. 8).

On the other hand, if the object has already been registered at the click location (YES side of S23 in FIG. 8), the MPU
12 determines whether the current operation mode of the image connection program is the object execution mode (S2 in FIG. 8).
5). If in the object execution mode (FIG. 8S
(YES side of 25), the MPU 12 generates a child thread for executing the registered object. This child thread executes the registered object in synchronization with the scroll operation (steps S18 to S21 in FIG. 8) of the parent thread (S26 in FIG. 8).

On the other hand, in the case of the object editing mode (NO in S25 in FIG. 8), the MPU 12 displays a dialog for editing the object and accepts an instruction to edit the object (eg, editing, updating, deleting, etc.) from the operator. (FIG. 8S27). Through the above operation (steps S22 to S27 in FIG. 8), the operation of the object associated with the connected image can be performed.

(Window end instruction) When the window end message is received in the standby state (YES side in S28 in FIG. 8), the MPU 12 sequentially deletes each window and releases resources related to the image connection program. Execute (S29 in FIG. 8). After finishing such post-processing,
The MPU 12 exits from the message waiting state and ends the image connection program.

(Effects of Second Embodiment, etc.) With the operations described above, the second embodiment can provide the same effects as those of the first embodiment. Furthermore, in the second embodiment, a line segment in which the total area of unnecessary portions to be cut is as small as possible is selected as a cut line. Therefore, it is possible to minimize the disappearance of the image information after the connection processing.

In the second embodiment, a rotation process corresponding to the image display angle is executed during the image connection process.
Therefore, it is not necessary to perform the image rotation processing during the scroll display, and the speed of the scroll display can be further increased. Further, in the second embodiment,
Since the registered object can be executed by mouse click, an image connection system excellent in presenting various information is realized.

In the second embodiment, an object can be updated or newly registered by clicking the mouse, so that an image connection system excellent in flexible object operation can be realized. In particular, in the second embodiment, coordinate conversion parameters [NEW_θ ′, Q ′] that can execute coordinate conversion of a connected image in the shortest procedure are used. Therefore, it is possible to further increase the calculation efficiency when performing coordinate transformation of the connected image.

[0094]

【The invention's effect】

( Claim 1 ) In the invention according to claim 1 ,
Since a line segment that is as orthogonal as possible to the connecting direction of the images is selected as the cut line, it is possible to cut the connecting portions of a plurality of images in an appropriate direction. ( Claim 2 ) In the invention according to claim 2 , a line segment in which the total area of unnecessary portions to be cut is as small as possible is selected as a cut line. By setting such a cut line, it is possible to suppress the disappearance of the image information after the connection processing.

( Claim 3 ) In the invention according to claim 3 ,
Since the rotation processing of the image according to the image display angle is completed during the image connection, the rotation processing of the image during the scroll display can be omitted. Therefore, the scroll display speed can be further increased. ( Claim 4 ) In the invention according to claim 4 , a registered object in the object-oriented database is selected and executed in response to an external position instruction, so that an image connection system excellent in presenting various information is realized.

( Claim 5 ) In the invention according to claim 5 ,
Since the object is updated or newly registered in the object-oriented database in response to an external position instruction, an image connection system excellent in flexible information operation is realized. ( Claim 6 ) The computer-readable storage medium according to any one of claims 1 to 2 , using the recording medium described in claim 6.
The image connection system described in the section can be realized.

( Claim 7 ) The use of the recording medium according to claim 7 allows the computer to execute the method according to claims 3 to 5 on a computer.
The image connection system according to any one of the above items can be realized.

Claims (11)

[Claims] 1. An image connection system for connecting a plurality of images, wherein the plurality of images are displayed together or sequentially to receive a position input on a screen, and a corresponding point overlapping between the plurality of images is detected. A position input unit that captures at least two places, a movement amount calculation unit that calculates a translation amount and a rotation amount for overlapping a corresponding point of another image on a corresponding point of one image obtained in the position input unit, An image arranging unit that performs coordinate conversion of the other images using the parallel movement amount and the rotation amount calculated by the movement amount calculation unit, and arranges the plurality of images on a common coordinate space. Image connection system. 2. The image connection system according to claim 1, wherein the movement amount calculation unit determines a vector connecting one corresponding point and another corresponding point in each image, and A rotation amount calculation unit that calculates a relative angle of the vector between the images and calculates the rotation amount from the relative angle; calculates a distance between the one corresponding point between the plurality of images; and calculates the parallel distance from the distance. An image connection system, comprising: a parallel movement amount calculation unit for obtaining a movement amount. 3. The image connection system according to claim 1, wherein the movement amount calculation unit calculates an interval between corresponding points in each image, and calculates a distance between the plurality of images based on a ratio of the interval. The image placement unit has a scale calculation unit that calculates a scale relationship, and the image placement unit enlarges or reduces an image based on the scale relationship obtained by the scale calculation unit, and calculates a scale relationship of an image to be arranged on the common coordinate space. An image connection system characterized by matching. 4. The image connection system according to claim 1, wherein the image arrangement unit arranges each image on the common coordinate space by using the position input unit. An image connection system characterized in that unnecessary portions of each image are cut by line segments passing through corresponding points. 5. The image connecting system according to claim 4, wherein the image arranging unit determines, from a plurality of line segments passing through two or more corresponding points, a line segment having an angle close to a right angle with the image connecting direction. An image linking system characterized by cutting out unnecessary portions of each image using selected line segments. 6. The image connection system according to claim 4, wherein the image arranging unit selects, from a plurality of line segments passing through two or more corresponding points, a line segment having the smallest total area of unnecessary portions to be cut. Pick out,
An image connection system characterized in that unnecessary portions of each image are cut using selected line segments. 7. The image connection system according to claim 1, wherein a plurality of images arranged in the common coordinate space are rotated according to an externally input image display angle. An image storage unit that processes and stores the connected images as a connected image; and a scroll display unit that sequentially displays the connected images stored in the image storage unit in accordance with a scroll direction and performs a scroll display across a plurality of connected images. An image connection system, comprising: completing rotation processing of an image according to the image display angle during image connection so that rotation processing of an image during scroll display is not required. 8. The image connection system according to claim 7, wherein the image storage unit associates at least one of moving image information, still image information, audio information, command information, or character information with the stored connected image. An object-oriented database in which objects including the two can be registered, and an object selected from the object-oriented database in response to a position instruction on the screen given from the outside, and the selected object is executed according to the image display timing An image connection system comprising an object execution unit. 9. The image connection system according to claim 7, wherein the image storage unit associates at least one of moving image information, still image information, audio information, command information, or character information with the stored connected image. An object-oriented database that can register objects including at least one of the following, and an object that updates the object-oriented database by executing at least one of the following (1) and (2) in response to an externally given position indication on the screen: An image connection system comprising: an update unit. (1) a process of selecting a registered object corresponding to the position indication from the object-oriented database and executing any one of editing, updating, and deletion on the selected registered object (2) processing corresponding to the position indication Processing for newly registering an object in the object-oriented database 10. A machine-readable computer that stores an image connection program for causing a computer to function as the position input unit, the movement amount calculation unit, and the image arrangement unit according to any one of claims 1 to 6. recoding media. 11. An image connection for causing a computer to function as the position input unit, the movement amount calculation unit, the image arrangement unit, the image storage unit, and the scroll display unit according to any one of claims 7 to 9. A machine-readable recording medium that records a program.