JP2000207570A - Device and method for extracting contour line and storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To stably extract a contour line with high precision, which is adaptive to a wide range of images. SOLUTION: An image converting device 12 forms plural conversion images being different in resolution from an input image from an image signal input terminal 10. An image display device 14 simultaneously displays the plural conversion images which are obtained from the device 12. A contour extracting section designating device 18 designates a section for extracting the contour line. A contour line extracting device 16 extracts the contour line in the section which is designated by the device 18.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and a method for extracting contours and a storage medium, and more particularly, to extracting an object from an image and performing geometrical operations such as moving, inverting, rotating, scaling and deformation. Image editing and processing tool for performing various processing, image synthesis tool for performing inlay synthesis on other images, image processing tool for performing image processing such as color conversion or filtering processing only in a specified area, and image processing The present invention relates to an apparatus and method for extracting contours used in computer vision and the like that understands a scene by dividing an image into regions, and a storage medium.

[0002]

2. Description of the Related Art A method of extracting a contour line of an object from an image is classified into the following three types. That is, the first method is disclosed in JP-B-63-53587 and JP-B-63-55.
No. 101 and U.S. Pat. No. 5,471,535 (JP-A-5-81429),
In this method, continuous boundary lines are extracted by directly using the luminance level distribution of the input image or each color image (for example, each of R, G, and B images) obtained by color separation of the input image. As described in U.S. Pat. No. 5,617,487 (Japanese Unexamined Patent Application Publication No. 7-92650), a second method is to convert an input color image (RGB image or YMCK image) into a hue image H and a saturation image. This is a method of generating images of different color representation methods, S and brightness image I, and searching for a contour line using a total of six image data. A third method is to decompose the input image into several different resolution levels to form a pyramid structure (hierarchical structure), obtain coarse outline positions at a low resolution level, and shift to a high resolution level, This is a method of gradually increasing the contour position accuracy, and is used together with the first method or the second method. For example, Fujimura, Yokoya and Yamamoto, "Contour Tracking and Motion Analysis of Non-rigid Object Using Active Contour Model Using Multiscale Image" (Transactions of IEICE D-II, Vol. J76-D-II, No. .2,
pp. 382-390, February 1993).

[0003]

However, in the first conventional method, a black-and-white image having only luminance information or a single-color image having only one color information is used. Is required to have a sufficient contrast or color difference. Therefore, the first method can be applied only to a relatively simple image in which a subject is clearly seen in a flat background portion, for example.

A second method is to use an R image, a G image, a B image,
Since the result of the highest contrast can be used by using the six image data of the H image, the S image, and the I image, it is possible to deal with a wide variety of images as compared with the first method. However, it is difficult to apply the method to an actual image because it is easily affected by background noise and a fine pattern near an outline to be extracted.

A third method is a hierarchical image set having a pyramid structure, wherein processing at a certain hierarchical level affects processing at a higher hierarchical level or a lower hierarchical level. Therefore, if contour extraction fails at the preceding level, it will break down thereafter. In this regard, the third method is not particularly suitable for extracting low-contrast contour lines.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a contour extracting apparatus and method which can be applied to a wide range of images and can extract contours with high accuracy and stability, and a storage medium.

[0007]

SUMMARY OF THE INVENTION A contour line extracting device according to the present invention is a contour line extracting device for extracting a contour line from an input image, wherein the image extracting device generates a plurality of transformed images having different properties from the input image. Converting means, image display means for displaying at least one image of the input image and the plurality of converted images, contour extraction section designating means for specifying a contour extraction section to be extracted from the input image; In each of the contour extraction sections designated by the contour extraction section designation means, a contour line extraction means for extracting a contour line from one or more of the plurality of converted images is provided.

A contour line extracting method according to the present invention includes an image converting step of generating a plurality of converted images having different properties from an input image; a contour extracting section specifying step of specifying a contour extracting section to be extracted from the input image; A contour extraction step of extracting a contour line in the contour extraction section specified in the contour extraction section specifying step, and a conversion image suitable for contour extraction of the contour extraction section specified in the contour extraction section specifying step. A step of selecting and confirming the extracted contour line extraction result as a contour line.

[0009] A contour line extracting method according to the present invention is also a contour line extracting method for extracting a contour line from an input image, wherein an image converting step of generating a plurality of transformed images having different properties from the input image; An image display step of displaying an input image and at least one image of the plurality of converted images, a contour extraction section designating step of designating a contour extraction section to be extracted from the input image, and a contour extraction section designating step A contour extraction step of extracting a contour from one or more of the plurality of converted images in each contour extraction section specified by the following.

[0010] The storage medium according to the present invention stores the program and software of the above-described contour line extraction method.

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

FIG. 1 is a structural block diagram showing an embodiment of the present invention. 10 is an image signal input terminal, 12 is an image conversion device, 14 is an image display device, 16 is a contour line extraction device, 1
8 is a contour extraction section designating device for designating a contour extraction section, 2
0 is a contour address output terminal for outputting an address indicating the extracted contour. In the present embodiment, not only hardware but also a part or all thereof can be realized by software.

The purpose of the image converter 12 is to make it easier to extract contour lines from an input image. FIG. 2 shows a configuration example of the image conversion device 12. The image signal from the image signal input terminal 10 is supplied from the image signal input terminal 22 to the image processing circuits 24-1 to 24-4.

The image processing circuit 24-1 converts the input image into a first image.
Smoothing is performed by the following method. The smoothing process # 1 includes, for example,
This is a filtering process (convolution operation) by performing a product-sum operation on a mask having a size of 3 × 3 pixels as shown in FIGS. Edge extraction processing circuit 26-1
Extracts an edge from the image signal smoothed by the image processing circuit 24-1. The edge extraction is realized by, for example, a filtering process by a well-known Sobel operator. The converted image # 1 output from the edge extraction processing circuit 26-1 is output from the converted image output terminal 28-1 to the outside.

The image processing circuit 24-2 smoothes the image signal from the input terminal 22 by the second method. The smoothing process # 2 in the image processing circuit 24-2 uses, for example, a mask having a size of 5 × 5 pixels as shown in FIGS. 5 and 6, and is stronger than the smoothing process # 1. Become The edge extraction processing circuit 26-2 extracts an edge by the same processing as the edge extraction processing circuit 26-1. The edge extraction processing circuit 26-2 may perform edge extraction in accordance with strong smoothing in the image processing circuit 24-2. The converted image # 2 output from the edge extraction processing circuit 26-2 is output from the converted image output terminal 28-2 to the outside.

In this embodiment, there are two levels of smoothing. By performing the N-stage smoothing process, N levels having different resolutions are obtained.
Needless to say, one converted image can be obtained. The number N of smoothing levels can be appropriately set as needed. By generating a plurality of converted images of different resolutions and selecting a desired one among them, for example, selecting a high-resolution converted image for extracting a contour line from a low-contrast image, When extracting a contour from an image having much noise near the line, a low-resolution converted image can be selected, and the contour can be correctly extracted from a wider image.

The image processing circuit 24-3 generates a color difference image # 1 from the image from the input terminal 22. The definition of the color difference image # 1 is shown in the following equation. That is, color difference image # 1 = (R image) -one (B image) (1) However, the input images are an R (red) image, a G (green) image, and a B (green) image.
(Blue) It is assumed that the image is a color image composed of three color images.

The edge extraction processing circuit 26-3 may perform the same processing as that performed by the edge extraction processing circuit 26-1 on the color difference image # 1, or may perform edge extraction processing having characteristics adapted to the color difference image # 1. You may do it. The converted image # 3 thus generated is output from the converted image output terminal 28-3.

The image processing circuit 24-4 generates a color difference image # 2 defined by the following equation from the input image. That is, color difference image # 2 = 2 × (G image) − (R image) − (B image) (2) However, also in this case, the input image is an R (red) image, G (green)
An image and a B (blue) image are three-color images.

The edge extraction processing circuit 26-4 may perform the same processing as that of the edge extraction processing circuit 26-1 on the color difference image # 2, or may perform edge extraction processing having characteristics matching the color difference image # 2. May be. The converted image # 4 thus generated is output from the converted image output terminal 28-4.

The operation of the embodiment shown in FIG. 1 will be described. The image conversion device 12 performs some different image processing on the image signal input from the image signal input terminal 10 as described with reference to FIG. The purpose of image processing in the image conversion device 12 is to make it easier to extract a contour from an input image.

In the present embodiment, the above-described image conversion is applied to the entire surface of the input image. However, in order to increase the speed, the image is limited to the area designated by the contour extraction section designating device 18 or its vicinity. A transformation may be applied. Further, the color image is not limited to the RGB format. CMY model or YI
It may be based on a Q model or the like. The definition of the color difference image is not limited to the above equations (1) and (2), and other known color difference images or hue images H may be used.

The converted image obtained by the image conversion device 12 is displayed on the screen of the image display device 14. FIG.
4 output from the converted image output terminals 28-1 to 28-4
5 shows a display example of an image display device 14 of a number of converted images.
1 shows a polybucket® placed on a carpet. Four converted images 30 displayed on the sub-screen of FIG.
32, 34, and 36 are obtained through partially modified image processing contents described with reference to FIG.

The converted image 30 is an edge image obtained by directly applying a sobel operator to an input image.
The converted image 32 is an edge image obtained by applying the mask shown in FIG. 3 to the input image, smoothing the input image, and then applying the Sobel operator. The converted image 34 is an edge image obtained by applying the mask shown in FIG. 5 to the input image, smoothing the image, and then applying the Sobel operator. The converted image 36 is smoothed by applying the mask shown in FIG. 5 to the color difference image # 1 defined by the equation (1),
It is a color difference edge image obtained by applying an operator.
Therefore, FIG. 7 shows three images having different resolutions and four converted images obtained by extracting edges from an image obtained by performing a smoothing process on the color difference image # 1 defined by the equation (1). .

In this embodiment, the user can extract a contour line using an optimum converted image for each contour section to be extracted. In other words, the user can display 4 displayed as shown in FIG.
From the converted images, a converted image in which the currently focused contour section is the clearest is selected. Then, a contour section to be extracted is designated by the contour extraction section designation device 18 on the selected converted image. As a specific method of specifying the contour section, for example, a method of specifying both end points of the contour or a plurality of passing points on the contour with a pointing device such as a mouse, or a method of roughly tracing the contour may be used.

The contour extraction unit 16 extracts a contour in a section designated by the contour extraction section designation unit 18 by a known method. The contour lines and their candidates obtained by the contour line extraction device 16 are equally superimposed and displayed on all the converted images shown in FIG. 7 as exemplified by reference numerals 38, 40, and 42. The contour lines 38 and 40 are extracted from the converted images 32 and 34, respectively. The outline candidate 42 indicated by the broken line is extracted from the converted image 30. If the contour candidate 42 is correct, the user
Specify a new contour line extraction section. By designating the next contour extraction section, the contour candidate 42 is automatically determined as a contour, and the line type or color is displayed so as to indicate that the contour is the determined contour in FIG. The method is changed. In FIG. 7, the contour line candidates 42 before determination are displayed by broken lines, and the contour lines 38 and 40 after determination are displayed by solid lines.

The user selects a converted image whose contour section of interest is the sharpest, and specifies a contour section to be extracted on the selected converted image by the contour extraction section specifying device 18. Since the contour line and the contour line candidates determined by the operations up to now are displayed on all the converted images, the progress of the contour line extraction work is the same on any of the converted images, which is obvious at a glance. Therefore, the user can obtain all the address values of the outline by simply selecting the converted image having the sharpest outline for each outline section of interest and repeatedly specifying the outline extraction section. The address of the contour obtained in this manner is output from the contour address output terminal 20.

If an accurate contour cannot be extracted even if any converted image is used and the contour extraction section is specified again, the contour is input freehand, and the contour extraction section is connected by a straight line. Then, a mode other than the contour line extraction mode, such as performing a smoothing process on the contour line candidate, is activated.

The present embodiment has the following specific effects. In other words, the user can select the image with the sharpest outline to be extracted while comparing a plurality of converted images having different characteristics, so that the conventional technology can extract an image including a complicated background or an image with low contrast. A difficult contour can be easily and accurately extracted. In addition, by displaying the determined outline and the outline candidate equally on all the converted images, the operation of switching a plurality of converted images for each outline part from which extraction is desired can be smoothly performed without being confused.

In the above embodiment, the contour was extracted using one optimal converted image determined by the user. However, contour extraction was performed on all the converted images in parallel, and each contour line candidate was converted to the corresponding converted image. You may superimpose and display on an image. In this case, immediately after contour extraction, a result of contour extraction (contour candidate) using the converted image is displayed on each converted image. Here, the user selects and determines the contour candidate from which the contour is most accurately extracted. The contour after confirmation is
As described above, the image is superimposed and displayed equally on all the converted images. If the correct outline is not included in the plurality of outline candidates, the outline extraction section is specified again. If an accurate contour cannot be extracted even if the contour extraction section is changed, input the contour by freehand, connect the contour extraction section with a straight line, and perform smoothing processing on the contour candidate. Activate a mode other than the line extraction mode.

Such a change has the following specific effects in addition to the effects described above. In other words, only by specifying the contour extraction section once, the contour line candidates corresponding to the number of converted images are displayed. Since only an operation of selecting a correct contour line from those contour line candidates is sufficient, a correct contour line can be obtained more quickly.

Although a plurality of converted images are displayed simultaneously, one converted image or only an input image may be displayed. Even in that case, internally, the image conversion device 12 generates and holds a plurality of converted images. The user designates a section to be subjected to contour extraction by the contour extraction section designation device 18 while viewing the displayed one image. The contour line extraction device 16 extracts a contour line using each of the plurality of converted images independently.
The result of the extraction of a plurality of contours, that is, 1
One outline candidate is selected and superimposed on the image being displayed, and waits for the user's judgment. The order in which the contour line candidates are superimposed and displayed may be in accordance with the order of the converted images, or may be in accordance with the order of contour line candidates that are assumed to be more correct. As the criterion for determining the likelihood of the contour, the smoothness and length of the contour, and the average and variance of the edge strength on the contour can be used.

If the displayed contour line candidate is correct, the user specifies the next contour extraction section. By the operation of specifying the contour extraction section, the contour line candidate is determined as a contour line, and is changed to a display mode so as to be distinguishable from the contour line candidate. If the displayed contour candidate is incorrect, the user sequentially displays other contour extraction results using the other converted images, and selects a correct contour. If the correct contour is not included in the contour candidates, the contour extraction section is specified again. If an accurate contour cannot be extracted even if the contour extraction section is changed, a contour such as inputting a freehand contour, connecting the contour extraction section with a straight line, and performing a smoothing process on the contour candidate is used. Start a mode other than the extraction mode.

FIG. 8 is a schematic diagram showing a state in which only one converted image 44 is displayed as a representative of a plurality of converted images, and an outline is extracted on the converted image 44. The outlines 46 and 48 drawn by solid lines have already been determined. The outline 46 is extracted from the converted image 30, and the outline 48 is extracted from the converted image 32. The dashed outline candidate 50 is extracted from the converted image 30 and indicates a correct outline. Contour candidate 50
Is automatically determined as a contour as soon as the next contour extraction operation is started, and the display form changes from a broken line to a solid line. Hereinafter, contour line extraction / confirmation work proceeds one after another in a similar procedure.

The best contour extraction result can be selected by comparing contour candidates obtained by using a plurality of transformed images having different characteristics, so that it is difficult to extract an image including a complicated background or a low contrast image by the conventional technique. The contour line can be correctly extracted from the image. Further, since only one image is displayed, the configuration of the display screen is simplified, and the processing is facilitated.

A plurality of contour line candidates may be simultaneously superimposed on a single display image. FIG. 9 is a schematic diagram showing a result of extracting a contour by the contour extracting device 16 between both end points 52 and 54 of the contour specified by the contour extracting section specifying device 18. Reference numerals 56, 58, 60, and 62 indicate contour line candidates extracted using different converted images, respectively. The contour line candidates 56 to 62 are displayed so as to be easily distinguished by different display methods such as changing the line type or changing the color as shown in FIG. The user sees the screen illustrated in FIG. 9 and selects the most correct contour line candidate. When the correct outline is not included in the outline candidate,
What is necessary is just to change the outline extraction section. If an accurate contour cannot be extracted even if the contour extraction section is changed, a contour such as inputting a freehand contour, connecting a contour extraction section with a straight line, and performing a smoothing process on a contour candidate is used. Start a mode other than the extraction mode.

In this manner, by specifying the contour extraction section once, contour candidates are displayed by the number of converted images, and it is only necessary to select the correct contour from the contour candidates. The operation is simplified, and the burden on the user is greatly reduced. The contour that seems to be the most correct is automatically selected, and only the selection result is displayed. It is not possible to see the contour line candidates that are out of selection.

The present invention may be applied to a system constituted by a plurality of devices or to an apparatus constituted by a single device.

Further, in order to operate various devices so as to realize the functions of the above-described embodiment, software for realizing the functions of the above-described embodiment is installed in a computer in an apparatus or a system connected to the various devices. The present invention also includes a case in which a program (code) is supplied and a computer (CPU or MPU) of the apparatus or system is operated by operating the various devices according to stored programs.

In this case, the software program code itself implements the functions of the above-described embodiment, and the program code itself and means for supplying the program code to the computer, for example,
A storage medium storing such a program code constitutes the present invention. As a storage medium for storing such a program code, for example, a floppy (registered trademark) disk, hard disk, optical disk, magneto-optical disk, C
A D-ROM, a magnetic tape, a nonvolatile memory card, a ROM, and the like can be used.

When the computer executes the supplied program code, not only the functions of the above-described embodiment are realized, but also the OS (operating system) or other operating system running on the computer. Even when the functions of the above-described embodiments are realized in cooperation with application software and the like, it goes without saying that such program codes are included in the embodiments of the invention according to the present application.

Further, the supplied program code is
After being stored in the memory provided in the function expansion board of the computer or the function expansion unit connected to the computer, the CPU or the like provided in the function expansion board or the function expansion unit performs one of the actual processing based on the instruction of the program code. It is needless to say that a case where the functions of the above-described embodiments are realized by performing all or a part of the processes is also included in the scope of the invention according to the present application.

[0042]

As can be easily understood from the above description, according to the present invention, a plurality of transformed images having different properties are generated from an input image, and an optimal transformed image is selectively selected for each contour section to be extracted. , It is possible to accurately and stably extract an outline from a wide range of images.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram of an embodiment of the present invention.

FIG. 2 is a schematic block diagram of an image conversion device 12;

FIG. 3 is a diagram illustrating an example of a mask used for a smoothing process.

FIG. 4 is a diagram illustrating an example of a mask used for a smoothing process.

FIG. 5 is a diagram illustrating an example of a mask used for a smoothing process.

FIG. 6 is a diagram illustrating an example of a mask used for a smoothing process.

FIG. 7 is an example of a display screen.

FIG. 8 is another example of a display screen.

FIG. 9 is a schematic diagram of another example of a display screen.

[Explanation of symbols]

 10: Image signal input terminal 12: Image conversion device 14: Image display device 16: Outline extraction device 18: Outline extraction section designating device 20: Outline address output terminal 22: Image signal input terminal 24-1 to 24-4: Image processing circuits 26-1 to 26-4: Edge extraction processing circuits 28-1 to 28-4: Converted image output terminals 30, 32, 34, 36: Converted images 38, 40: Contour 42: Contour candidate 44: Converted images 46, 48: contour line 50: contour line candidate 52, 54: both end points of contour line 56, 58, 60, 62: contour line candidate

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Fumiaki Takahashi 3-30-2 Shimomaruko, Ota-ku, Tokyo F-term in Canon Inc. (reference) 5B057 CA01 CA02 CA08 CA12 CA16 CB01 CB02 CB08 CB12 CB16 CC02 CC03 CE05 CE06 CH09 DA20 DC16 5L096 AA01 AA02 AA06 CA24 EA06 FA06 GA10 GA38 GA55

Claims (19)

[Claims] 1. A contour line extracting device for extracting a contour line from an input image, comprising: image conversion means for generating a plurality of converted images having different properties from the input image; Image display means for displaying at least one of the images, contour extraction section designation means for designating a contour extraction section to be extracted from the input image, and contour extraction sections designated by the contour extraction section designation means. A contour line extracting unit for extracting a contour line from one or more of the plurality of converted images. 2. The contour line extracting device according to claim 1, wherein the plurality of converted images include a plurality of converted images having different resolutions. 3. The image display means displays the plurality of converted images simultaneously, and extracts from the one converted image selected from the plurality of converted images in the contour extraction section specified by the contour extraction section specifying means. The outline extracting apparatus according to claim 1, wherein the outline and the candidate for the outline are superimposed and displayed equally on all of the plurality of converted images. 4. The image display means displays the plurality of converted images at the same time, and a contour line executed in parallel with each of the plurality of converted images in a contour extraction section designated by the contour extraction section designating means. 2. The contour line extracting device according to claim 1, wherein a plurality of contour line candidates extracted by the extraction are superimposed and displayed on each of the converted images. 5. The image display means displays at least one of an input image and one converted image, and displays one of the plurality of converted images in a contour extraction section specified by the contour extraction section specifying means. The contour line extracting apparatus according to claim 1, wherein the contour line candidate extracted from the converted image is superimposed on the image being displayed. 6. The image display means displays at least one of an input image and one converted image, and displays a plurality of contour line candidates extracted using each of the plurality of converted images so as to be distinguishable from each other. 2. The contour line extracting device according to claim 1, wherein the contour line extracting device simultaneously superimposes and displays the image on the image. 7. The image display means displays at least one of an input image and one converted image, and extracts from a plurality of contour line extraction results extracted using each converted image of the plurality of converted images. The contour line extraction device according to claim 1, wherein a contour line candidate which seems to be the most contour is superimposed and displayed on the display image. 8. A graphical user displaying a plurality of converted images obtained by converting an input image, selecting an optimum converted image for each section of a contour to be extracted, and extracting a contour. ·interface. 9. A plurality of transformed images obtained by transforming an input image are displayed, each contour candidate extracted using each transformed image is superimposed on each corresponding transformed image, and a correct contour candidate is displayed. A graphical user interface characterized by selecting and confirming a contour line. 10. A graphical user interface wherein a plurality of contour line candidates of a section of interest whose contour is to be extracted are displayed so as to be distinguishable from each other, and a correct contour line candidate is selected and determined as a contour line. 11. An image conversion step for generating a plurality of converted images having different properties from an input image, a contour extraction section specifying step for specifying a contour extraction section to be extracted from the input image, and a contour extraction section specifying step A contour extraction step of extracting a contour in the extracted contour extraction section, and a contour extraction result extracted using a converted image suitable for contour extraction of the contour extraction section specified in the contour extraction section specifying step. A step of selecting and confirming a contour line. 12. An outline extraction method for extracting an outline from an input image, comprising: an image conversion step of generating a plurality of converted images having different properties from the input image; An image display step of displaying at least one of the images, a contour extraction section specifying step to specify a contour extraction section to be extracted from the input image, and a contour extraction section specified by the contour extraction section specifying step. A contour line extracting step of extracting a contour line from one or more of the plurality of converted images. 13. The contour extraction method according to claim 12, wherein the plurality of converted images include a plurality of converted images having different resolutions. 14. In the image display step, the plurality of converted images are displayed simultaneously, and one converted image selected from the plurality of converted images in the contour extraction section specified by the contour extraction section specifying step. The contour and the contour candidate extracted from are superimposed and displayed equally on all of the plurality of transformed images, and when it is determined that the contour candidate is correct, the process proceeds to a next contour extracting section designation operation, and the contour is extracted. When it is determined that the candidate is not correct, at least one of the outline extraction section and the converted image is changed to newly perform outline extraction, and the outline is gradually determined.
3. The contour extraction method according to 2. 15. In the image display step, the plurality of converted images are displayed simultaneously, and the plurality of converted images are simultaneously executed in the contour extraction section designated by the contour extraction section designation step. A plurality of contour line candidates extracted by contour line extraction are superimposed and displayed on each of the corresponding converted images. If the plurality of contour line candidates include a correct contour line, the correct contour line candidate is determined. After that, the process proceeds to the operation of specifying the next contour extraction section, and when the plurality of contour line candidates do not include a correct contour line, the contour extraction section is changed and a new contour line is extracted. 13. The contour extraction method according to claim 12, wherein the contour is gradually determined. 16. The image display step displays at least one of an input image and one converted image, and further includes displaying at least one of the plurality of converted images in the contour extraction section specified by the contour extraction section specifying step. The outline candidate extracted from one converted image is superimposed and displayed on the image being displayed, and when it is determined that the outline candidate is correct, the process proceeds to the operation of specifying the next outline extraction section, and the outline candidate is determined. When it is determined that the outline is not correct, the outline is gradually determined by repeating at least one of the outline extraction section and the converted image and performing a new outline extraction.
3. The contour extraction method according to 2. 17. In the image display step, at least one of an input image and one converted image is displayed, and a plurality of contour line candidates extracted using each of the plurality of converted images can be identified from each other. Is displayed simultaneously on the currently displayed image, and when the plurality of contour candidates include the correct contour, the contour candidate of the correct contour is determined, and then the operation proceeds to the operation of designating the next contour extracting section. If the plurality of contour candidates do not include a correct contour, the contour is gradually determined by re-designating a contour extraction section and performing new contour extraction. Described contour line extraction method. 18. The image display step includes displaying at least one of an input image and one converted image, and extracting a plurality of contour lines extracted using each of the converted images. From the result, a contour line candidate that seems to be the most contour is superimposed on the display image, and when it is determined that the contour line candidate is correct, the process proceeds to a next contour extraction section designating operation, and it is determined that the contour line candidate is not correct. 13. The outline extraction method according to claim 12, wherein when the outline is extracted, the outline is gradually determined by re-designating the outline extraction section and performing new outline extraction. 19. A storage medium storing program software for the contour line extraction method according to claim 11. Description: