JPH0981712A - Method and device for image processing - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable an optimum image segmentation processing adaptive to user's intention. SOLUTION: This method include a 1st selecting process (S201-S202) which selects a 2nd point closest to a 1st point out of pixels having density gradients larger than a specific threshold on a line that passes a 1st point on a specific line on an image and crosses the specific line almost at right angles and a 2nd selecting processes (S203) which selects a 4th point closest to the selected 2nd point out of the pixels having density gradients larger than the specific threshold nearby the 1st point that passes a 3rd point on the specific line and crosses the specific line almost at right angles; and the line passing the 2nd and 4th points selected by the 1st and 2nd selecting means is decided as an image segmentation line (S205).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method and an apparatus thereof, and more particularly to an image processing method and an apparatus thereof for cutting out a region from an image.

[0002]

2. Description of the Related Art Conventionally, as a method for cutting out a specific area of an image and removing the other area, for example, Japanese Patent Laid-Open No. 60-1
There was a method described in 69977. In this method, first, the user designates a representative point in the vicinity of the boundary line of the area to be cut out. Then, a point on the perpendicular of the line segment connecting the representative points and located within a certain fixed distance, where the differential value (edge) of each pixel value is the maximum, is selected. Further, by connecting these points, the boundary line for cutting out is determined.

In this method, further, the image is smoothed in advance, or the obtained points are fitted by the method of least squares to determine the boundary line, thereby eliminating the unevenness caused by the noise of the image. It was Further, when it is desired to blur the contour like the hair of a person, the surrounding pixel values are weighted and added to each pixel around the contour.

[0004]

However, such a conventional method has the following problems. That is,
In the vicinity of the point designated by the user, the pixel having the maximum differential value of the pixel value may not necessarily be the boundary point if the user wants to cut out the pixel. For example, when cutting out a region of a person from an image, for a region where many lines exist in the same direction like a wrinkle of clothes, the points where the differential value of the pixel value is the maximum are dispersed, As a result, cutting out is performed with a line having irregularities.

Further, the boundary has been cut out which is strongly affected by noise such as isolated points. Furthermore, it is impossible to change the blurring amount of a specific portion of the contour from that of other portions, and it is only possible to blur all contours uniformly. The present invention has been made in view of the above conventional example, and an object of the present invention is to provide an image processing method and an apparatus thereof capable of appropriately cutting out an image intended by a user.

[0006]

In order to achieve the above object, an image processing method and apparatus according to the present invention have the following arrangement. That is, the second pixel closest to the point on the line that passes through the first point on the predetermined line with respect to the image and is substantially orthogonal to the predetermined line and has a density gradient of a predetermined threshold value or more.
And a third threshold on the predetermined line in the vicinity of the first point, and on a line orthogonal to the predetermined line and having a predetermined threshold value or more. A second selection unit that selects a fourth point that is closest to the selected second point among pixels having a density gradient of 2 above, and is selected by the first and second selection units. The line that passes through the second point and the fourth point is the image cutout line.

According to another aspect of the invention, among the pixels having a density gradient of a predetermined threshold value or more on a line which passes through a first point on a predetermined line with respect to the image and is substantially orthogonal to the predetermined line, First selecting means for selecting the closest second point;
Of the pixels having a density gradient of a predetermined threshold value or more on a line passing through a third point on the predetermined line in the vicinity of the first point and orthogonal to the predetermined line, the selected pixel is selected. Second selecting means for selecting a fourth point closest to the second point, and the fourth point is not adjacent to another second or fourth point already selected, Invalidating means for invalidating the fourth point, wherein the selected valid second
And a line passing through the fourth point is set as an image cutout line.

Another aspect of the present invention is that, on a line that passes through a first point on a predetermined line with respect to an image and is substantially orthogonal to the predetermined line, the point having a density gradient of a predetermined threshold value or more is set to the point. A first selection step of selecting the closest second point;
Of the pixels having a density gradient of a predetermined threshold value or more on a line passing through a third point on the predetermined line in the vicinity of the first point and orthogonal to the predetermined line, the selected pixel is selected. A second selection step of selecting a fourth point closest to the second point, and a line passing through the second point and the fourth point selected in the first and second selection steps. Image cutout line.

According to another aspect of the invention, among the pixels having a density gradient of a predetermined threshold value or more on a line that passes through a first point on a predetermined line with respect to the image and is substantially orthogonal to the predetermined line, A first selection step of selecting the closest second point;
Of the pixels having a density gradient of a predetermined threshold value or more on a line passing through a third point on the predetermined line in the vicinity of the first point and orthogonal to the predetermined line, the selected pixel is selected. A second selection step of selecting a fourth point closest to the second point, and the fourth point is not adjacent to another second or fourth point already selected, A nullification step of nullifying the fourth point, wherein the selected valid second
And a line passing through the fourth point is set as an image cutout line.

Another invention is a computer program product, comprising a computer usable medium having computer readable program code means for cutting out a predetermined area from an image, said computer program product for an image. Computer readable to select a second point closest to the first point on a line that passes through the first point on the predetermined line and is substantially orthogonal to the predetermined line and has a density gradient of a predetermined threshold value or more. Na first
And a pixel having a density gradient of a predetermined threshold value or more on a line that passes through a third point on the predetermined line near the first point, passes through the program code means, and is orthogonal to the predetermined line. Computer-readable second program code means for selecting a fourth point closest to the selected second point, the second selected in the first and second selection steps. The line that passes through the point and the fourth point is the image cutout line.

Another aspect of the present invention is that, on a line which passes through a first point on a predetermined line with respect to an image and is substantially orthogonal to the predetermined line, the point having a density gradient of a predetermined threshold value or more is set to the point. A first selection step of selecting the closest second point;
Of the pixels having a density gradient of a predetermined threshold value or more on a line passing through a third point on the predetermined line in the vicinity of the first point and orthogonal to the predetermined line, the selected pixel is selected. A second selection step of selecting a fourth point closest to the second point, and the fourth point is not adjacent to another second or fourth point already selected, A nullification step of nullifying the fourth point, wherein the selected valid second
And a line passing through the fourth point is set as an image cutout line.

[0012]

BEST MODE FOR CARRYING OUT THE INVENTION First, a summary of the points in the embodiment of the present invention will be given, followed by a detailed description thereof. In the image area cutout processing of the present embodiment, first, a line is designated on the image. Then, with respect to each point on the line designated for determining the boundary point to be cut out, the distance is within a certain range, the feature amount is equal to or more than a predetermined value, and the boundary point is determined. Select the point closest to the point. Then, it is possible to cut out a desired image area by deleting the points that are not connected by a predetermined number or more from these points and interpolating the boundary points.

When the boundary points are interpolated, the lines designated on the image are first used to interpolate. Furthermore, the image is cut out along the cut-out boundary points, and the interpolated boundary points are blurred more strongly than the non-interpolated boundary points, thereby blurring the surroundings of the cut-out image. .

Further, when the boundary points of the cut-out area are changed and the interpolated boundary points are changed, the changed boundary points are blurred because they are not interpolated. Also,
With respect to the pixels in the vicinity of the boundary point, the average of the pixel values in the surroundings is used, and in order to increase the degree of blurring, more pixels are blurred or averaged with more pixels.

The points of the processing that characterizes the present invention have been summarized above. Next, these processes will be described in detail. FIG. 1 is a diagram showing a configuration example of processing of the present embodiment. An image storage unit 101 stores an original image and an image cut out from the original image.

A display unit 102 displays an image. An input unit 103 includes a pointing device for roughly specifying a cutout area in an image and a keyboard for inputting various commands. An image cutout unit 104 determines an image cutout area and actually cuts out the image.

Reference numeral 105 denotes a blurring processing unit for blurring the periphery of the cut out area. Reference numeral 106 denotes a correction unit for the user to correct the cut out area. Next, the image area cutout processing of the present embodiment will be described using the flowchart of FIG. Note that it is assumed that the image to be cut out is already displayed on the display unit 102.

First, in step S201, the user traces a region in the original image to be clipped with the input unit 103 with a roughly thick line. This line is called a user-specified line. The image cutout unit 104 inputs the locus of the user-specified line via the input unit 103 and displays the corresponding line on the display unit 102. The user-specified line to be displayed is shown in FIG.
As shown in, the wavy line (301) or the like is used to prevent the original image from disappearing as much as possible.

In step S202, the image cutout unit 1
04 determines, for each pixel on the center line of the user-specified line, in order to determine a differential value (hereinafter, referred to as an edge) of the pixel value with respect to the pixel existing in the area of the user-specified line,
A corresponding point on the boundary line (hereinafter referred to as a boundary point) is obtained. Specifically, the following processing is performed. FIG. 4 shows 3 of FIG.
The area 01 is enlarged and will be described with reference to this drawing.

In the figure, reference numeral 402 is the center line of the user-specified line, and the starting point of this line is 403. First, with respect to the point 403, the line 40 orthogonal to the traveling direction of the line 402
Each point in the user-specified line in the four directions is examined, and a point whose edge gradient is equal to or larger than a certain threshold and which is closest to the point 403 is selected. In FIG. 4, this point is indicated by 405, and this point is the starting point of the boundary line.

Then, a boundary point corresponding to another point on the center line of the user-specified line, for example, the point 406, is determined as follows. That is, the direction of the line 407 that is orthogonal to the traveling direction of the line 402 is checked, and among the points existing in the area of the user line, the edge is equal to or greater than a certain threshold and the boundary point selected immediately before (in this case, the point 408). ) Select the point closest to. In FIG. 4, the point 409 shows the selected point.

The reason why an edge having a gradient equal to or greater than a threshold value, which is an area within the user-specified line, and which is closest to the immediately preceding boundary point is selected is as follows. First, a portion having only a small gradient edge equal to or less than a predetermined threshold is a featureless portion, and it is difficult for humans to determine where to cut out. Forcing a boundary point at such a point can lead to a false boundary.

Furthermore, a point having an edge with a large gradient equal to or greater than a certain threshold can be a candidate for a boundary point regardless of the magnitude of the value. The point closest to the immediately preceding boundary point is selected in order to make the boundary line smoother. For example, in FIG. 5, when the boundary point for the point 601 located on the center line of the user-specified line is obtained, even if the gradient of the edge point at the point 602 is slightly higher than 603, if both values are equal to or greater than the threshold value. For example, the point 603 closer to the border point immediately before is selected.

Next, in step S204, the image cutout unit leaves only a predetermined number or more, for example, three or more, of the boundary points obtained as described above, and deletes other boundary points. This process will be specifically described below. For example, it is assumed that the boundary points are extracted as shown in FIG. 6 in step S203. FIG. 6 shows a state where four connected boundary points 501, one boundary point 502, and three connected boundary points (503) are extracted. A location where there is one isolated boundary point 502 is, for example, a portion having an edge such as hair, and the boundary point 5
It is highly possible that 01 and the boundary point 503 have one line and the boundary point 502 has another line. In such a state, no matter which line is selected as the boundary line, there is no great difference in the shape of the region after cutting out, and it looks natural to the user, but on the contrary, two lines are alternately selected, that is, If the boundary points 501, 502, and 503 are cut out by a boundary line that connects them, unevenness of the region occurs, and the image looks unnatural to the user. To prevent this, the boundary points 502 that are isolated
Can be deleted, which makes it possible to cut out a smooth area that is close to the shape of the area desired by the user and has no unevenness.

Next, in step 205, the image clipping unit 104 creates a boundary line based on the boundary points obtained up to step S204. This operation is performed by connecting the boundary points with a straight line if they are adjacent to each other. However, if there is no point having an edge with a gradient equal to or greater than the threshold value in step S203, or step S20.
In 4, the boundary points are not adjacent to the positions where the boundary points are deleted because the predetermined number or more of points are not connected. In that case, a portion where no boundary point exists is interpolated by the following method.

The first method is a method of performing interpolation using front and rear boundary points with a Bezier curve or a straight line. Second
Is a method of using the curved shape of the user-specified line, which will be described with reference to FIG. In FIG. 8, 801 and 802 are boundary points, and 803 and 804 indicate both ends of the user-specified line. Here, first, a section of the user-specified line designated by the line 805 passing through the boundary point 801 and orthogonal to the user-specified line and the line 806 passing through the boundary point 802 and orthogonal to the user-specified line is obtained. Then, the line segment in the section of the user-specified line is fitted between the sections of the boundary points 801 and 802, and connected as a line 807 to complete the interpolation.

After the boundary line is created by the above method, the boundary line is displayed overlaid on the original image. If the user looks at the displayed boundary line and the result is as expected, the user further extends the boundary line by executing steps S201 to S205. Then, the same step can be executed to start over.

Then, the same processing is repeated until the boundary line created in the above steps forms a desired area.
In step 206, a predetermined determination command is input from the input unit 103 to determine whether or not the clipped area is determined by the above processing. Then, if a decision command is input, the process proceeds to step S207, and if not input, the process returns to step S201 and repeats the same processing.

In step 207, the image cutout unit 10
In 4, the area is cut out using the determined boundary line,
The cut-out area is stored in the image storage unit 101, and the boundary line is displayed on the display unit 102. FIG. 7 shows an example in which a region is cut out and displayed on the image of FIG. that's all,
The processing operation for cutting out a desired area from the original image has been described. However, after cutting out the area in step S207 of FIG. 2, the blurring unit 105 may blur the periphery of the image area to be cut out before displaying the area. it can. As a result, it is possible to give a natural impression to the human appearance and a visual effect similar to soft focus. Next, this method will be described.

When the pixel value of the pixel (i, j) is x (i, j), the pixel (i ', j') near the boundary point to be drawn

[0031]

[Equation 1]

Xh is calculated. Here, m is a constant and determines the number of pixels of the pixel (i, j). This m
The larger the value, the larger the blur amount. Here, m
The degree of blurring can be changed by making the variable according to the boundary points. Specifically, in step S205, when a boundary line is created from the boundary points, interpolation is performed for boundary points that are not adjacent as described above, but in the vicinity of the boundary points that have undergone such interpolation. The degree of blurring is increased for the pixels. That is, the value of m is increased and the average with many pixels is calculated. Or blur only the interpolated part,
Do not blur other parts. The location where interpolation is performed is a location where the gradient value of the edge is small, or where there are many edges with a small gradient. By blurring such a location, the pseudo boundary line created by interpolation is created. You can create vague and more natural boundaries.

Further, not only the blur amount but also the blur width can be made variable. The blur width is the number of pixels (i ', j') in the vicinity taken for each boundary point. This number,
The blur width is increased by increasing the number of boundary points created by interpolation. As an example, FIG. 9 shows a state in which the image of FIG. 7 is blurred. Further, the user can correct the area cut out by the above result.

This processing is executed by the correction unit 106. This processing will be described with reference to FIG. FIG.
Indicates that a part of the image area is corrected and enlarged. In FIG. 10, reference numeral 1001 denotes a tool operated by the pointing device of the input unit 103. With this tool, the contour line is deformed by pushing or pulling the cut-out contour line determined in the above processing.

The tool 1001 moves so as to move in accordance with the movement of the pointing device 204.
The movement is controlled by 06. For example, the wavy line 1 in FIG.
It is assumed that 002 is the cut-out contour line determined by the above processing. In this state, the user cuts out the contour line 10
When it is desired to shift the position of 02, the tool 1001 is moved to the position of the cut-out contour line 1002 to be moved and pushed outward, so that the cut-out contour line 1002 is moved and deformed outward at the position.

The correction section 106 also includes the image storage section 101.
The original image can be read out from and the original image corresponding to the moved and deformed portion can be enlarged and displayed. Further, the blurring unit 105 also performs the same blurring processing on the corrected area as described above. Also in this case, if the corrected portion is the boundary line created by the interpolation in step S205, the blurring amount and blurring width equivalent to those of the non-interpolated portion are blurred. Such a boundary line should be strongly blurred as described above, but the new boundary line created by the user's positive correction accurately reflects the cutout area desired by the user, It is not necessary to increase the blur amount and blur width.

It is also possible to erase a part of the cut out area by using the tool 1001. This is done, for example, by pressing a predetermined key on the keyboard of the input unit 103 to switch to the erasing mode and designating that a predetermined portion of the cut out area is surrounded by the tool 1001. In this case, that portion will be erased. In this case as well, as in the previous case, if the boundary is created by interpolation, blurring equivalent to that of the non-interpolated portion is performed.

Next, a hardware configuration capable of implementing the image processing of this embodiment will be described with reference to FIG. The CPU 203 controls the entire image processing apparatus according to the present embodiment by a program stored in the memory 202 and corresponding to the above-described flowchart in FIG. 2 (corresponding to the blurring unit 105, the correction unit 106, and the image cutout unit 104 in FIG. 1). Various image processing programs (including programs) are read, interpreted, and executed.

The keyboard 203 and pointing device 204 correspond to the input unit 103 in FIG. Memory 2
Reference numeral 02 denotes an RO that incorporates the above-mentioned image processing programs
An M area and a RAM area (including an area corresponding to the image storage unit 101 in FIG. 1) used as various work areas are provided.

Next, FIG. 12 shows an example of the layout of each program corresponding to each process of the above-mentioned flowchart assigned to the ROM area of the memory 202. still,
It goes without saying that this program may be stored in a portable medium such as a floppy disk, loaded into the memory 202 at the time of execution, and executed by the CPU 200.

A line input program 2000 corresponds to the processing in step S201. Reference numeral 2001 is an edge extraction program, which corresponds to the processing in step S202. 2002 is an edge determination program,
It corresponds to steps S203 and S204.

2003 is a program corresponding to the above-mentioned line interpolation processing. Reference numeral 2004 is a program for performing the smoothing processing around the cutout contour described above. 2005 is
This is a program for extracting and displaying the cut-out area that has been smoothed. The present invention may be applied to a system including a plurality of devices such as a host computer, an interface, and a printer, or may be applied to an apparatus including a single device such as a copying machine. Needless to say, the present invention can be applied to a case where the present invention is achieved by supplying a program to a system or an apparatus. In this case, the system or device enjoys the effects of the present invention by reading the program into the system or device from a storage medium storing a program represented by software for achieving the present invention. It becomes possible.

As described above, according to the present embodiment, in order to determine the boundary point for cutting out the image, the feature amount that is at a certain distance from the point traced by the user and that is a certain amount or more. Since it checks whether it has (edge gradient), it does not select the boundary point at the place where the feature amount is less than a certain amount in the image and it is not known where to set the boundary. Reduces the chance of picking unintended boundary points.

Therefore, as compared with other methods for determining the boundary points, for example, the method for selecting the point having the maximum feature amount, a more accurate and less uneven surface is cut out. Become. Further, since the point closest to the point once determined as the boundary point is selected, the boundary points are selected adjacent to each other, and a smooth area with less unevenness is cut out. .

Since a certain number of points that are not connected are deleted from the boundary points, there is no unevenness in the cutout area that occurs when connecting isolated connecting points. According to the second embodiment, when the boundary points are not adjacent to each other, the boundary points are interpolated based on the line traced by the user. Therefore, even if the boundary points are distant from each other, a straight line or It is possible to cut out a region in a shape desired by the user as compared with interpolation using a Bezier curve or the like.

Further, for the boundary points on which the above-mentioned interpolation has been performed, the degree of blurring is made stronger than the boundary points on which the above interpolation has not been performed, so that the periphery of the clipped image is blurred, so pseudo-interpolation is performed. It is possible to make the boundary line ambiguous and to cut out the region in a shape desired by the user. In addition, it is possible to change the boundary line, and if the boundary point that has been interpolated is changed, the boundary point after the change will be blurred as if no work was performed. By blurring the boundary line when the user positively changes and the ambiguity of the boundary point is eliminated,
It is possible to cut out the region in a shape desired by the user.

[0047]

As described above, according to the present invention, it is possible to appropriately cut out an image intended by the user.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration concept of image processing according to an embodiment of the present invention.

FIG. 2 is a flowchart showing a procedure for cutting out an image area.

FIG. 3 is a diagram showing an area designated by a user.

FIG. 4 is a diagram for explaining a method of selecting a boundary point.

FIG. 5 is a diagram showing how a point close to a boundary point immediately before is selected.

FIG. 6 is a diagram showing a state of an isolated boundary point.

FIG. 7 is a diagram illustrating an example of a clipped image.

FIG. 8 is a diagram for explaining how boundary points are interpolated.

FIG. 9 is a diagram illustrating a state where a boundary line is blurred.

FIG. 10 is a diagram showing how an area is modified.

FIG. 11 is a diagram of a hardware configuration of the image processing apparatus according to the present embodiment.

FIG. 12 is a diagram showing a layout configuration of a memory of the image processing apparatus according to the present embodiment.

[Explanation of symbols]

 101 image storage unit 102 display unit 103 input unit 104 image cutout unit 105 blurring unit 106 correction unit

Claims (39)

[Claims] 1. A second pixel, which is on a line that passes through a first point on a predetermined line with respect to an image and is substantially orthogonal to the predetermined line, and is closest to the point among pixels having a density gradient of a predetermined threshold value or more. First selection means for selecting a point and a third point on the predetermined line in the vicinity of the first point and on a line orthogonal to the predetermined line and having a predetermined threshold value or more. A second selecting unit that selects a fourth point closest to the selected second point among pixels having a density gradient, and the first and second selecting units select the fourth point that is selected by the first and second selecting units. An image processing apparatus, wherein a line passing through the second point and the fourth point is used as an image cutout line. 2. The input device according to claim 1, further comprising an input unit for inputting a band-shaped region position including a desired image cutout line and setting a center line of the band-shaped region as a predetermined line for the image. Image processing device. 3. In the line passing through the selected second point and fourth point, there is no pixel having a density gradient greater than or equal to the predetermined threshold value in the second selection means, and the fourth point is The method further comprises line interpolation means for performing line interpolation of the unselected portion, wherein the line interpolation means is located at the fourth point or the second point in both directions with respect to the portion where the fourth point is not selected. From the two nearest points, and a line segment between the intersections of two lines that pass through each of the two points and are substantially orthogonal to the center line and a line that surrounds the band-shaped region is defined between the two points. The image processing apparatus according to claim 2, wherein the image processing apparatus is inserted. 4. The smoothing means for causing a predetermined smoothing filter to act on the periphery passing through the selected second point and fourth point, as claimed in any one of claims 1 to 3.
An image processing apparatus according to any one of claims 1 to 4. 5. The image processing apparatus according to claim 4, further comprising display means for displaying the image smoothed by the smoothing means. 6. The image processing apparatus according to claim 4, wherein the smoothing unit smoothes the periphery of the line interpolated by the interpolation unit more strongly. 7. A second pixel, which is on a line that passes through a first point on a predetermined line with respect to the image and is substantially orthogonal to the predetermined line, and is closest to the point among pixels having a density gradient of a predetermined threshold value or more. First selection means for selecting a point and a third point on the predetermined line in the vicinity of the first point and on a line orthogonal to the predetermined line and having a predetermined threshold value or more. Second selecting means for selecting a fourth point closest to the selected second point among pixels having a density gradient; and a predetermined number or more of the fourth points already selected. And a disabling means for disabling the fourth point when it is not adjacent to the second or fourth point of the image, and a line passing through the selected valid second and fourth points is cut out from the image. An image processing device characterized by being a line. 8. The input device according to claim 7, further comprising input means for inputting a band-shaped region position including a desired image cutout line and setting a center line of the band-shaped region as a predetermined line for the image. Image processing device. 9. In the line passing through the selected valid second and fourth points, there is no pixel having a density gradient equal to or higher than the predetermined threshold value in the second selection means, and the fourth point is The method further comprises line interpolation means for performing line interpolation of the unselected portion, wherein the line interpolation means is located at the fourth point or the second point in both directions with respect to the portion where the fourth point is not selected. From the two nearest points, and a line segment between the intersections of two lines that pass through each of the two points and are substantially orthogonal to the center line and a line that surrounds the band-shaped region is defined between the two points. The image processing device according to claim 8, wherein the image processing device is inserted. 10. The smoothing means for causing a predetermined smoothing filter to act on the periphery passing through the selected second point and fourth point, as claimed in any one of claims 1 to 9. The image processing device according to item 1. 11. The image processing apparatus according to claim 10, further comprising display means for displaying the image smoothed by the smoothing means. 12. The image processing apparatus according to claim 10, wherein the smoothing unit smoothes the periphery of the line interpolated by the interpolation unit more strongly. 13. A deformation means for deforming the image cutout line in the force direction by applying a predetermined virtual force to a predetermined position of the image cutout line. 12. The image processing device according to any one of 12. 14. A second pixel, which is on a line that passes through a first point on a predetermined line with respect to an image and is substantially orthogonal to the predetermined line, and is closest to the point among pixels having a density gradient of a predetermined threshold value or more. A first selection step of selecting a point, and a line passing through a third point on the predetermined line in the vicinity of the first point and orthogonal to the predetermined line and having a predetermined threshold value or more. A second selection step of selecting a fourth point closest to the selected second point among pixels having a density gradient, and a second selection step performed in the first and second selection steps. Point and fourth
An image processing method, characterized in that a line passing through the points is used as an image cutout line. 15. The method according to claim 14, further comprising an input step of inputting a band-shaped region position including a desired image cutout line and setting a center line of the band-shaped region as a predetermined line for the image. Image processing method. 16. On a line passing through the selected second point and fourth point, there is no pixel having a density gradient equal to or higher than the predetermined threshold in the second selecting step, and the fourth point is The method further comprises a line interpolation step of performing line interpolation on an unselected portion, wherein the line interpolation step is performed on both sides of a fourth point or a second point with respect to a portion where the fourth point is not selected. From the two nearest points, and a line segment between the intersections of two lines that pass through each of the two points and are substantially orthogonal to the center line and a line that surrounds the band-shaped region is defined between the two points. The image processing method according to claim 15, further comprising: 17. The smoothing step of causing a predetermined smoothing filter to act on the periphery passing through the selected second point and fourth point, as claimed in any one of claims 14 to 16. The image processing method described in. 18. The image processing method according to claim 17, further comprising a step of displaying the image smoothed in the smoothing step. 19. The image processing method according to claim 17, wherein the smoothing step further strongly smoothes the periphery of the line interpolated in the interpolation step. 20. A second pixel, which is on a line that passes through a first point on a predetermined line with respect to an image and is substantially orthogonal to the predetermined line, and is closest to the point among pixels having a density gradient of a predetermined threshold value or more. A first selection step of selecting a point, and a line passing through a third point on the predetermined line in the vicinity of the first point and orthogonal to the predetermined line and having a predetermined threshold value or more. A second selecting step of selecting a fourth point closest to the selected second point among pixels having a density gradient; and a predetermined number or more of the fourth points having already been selected. A nullification step of nullifying the fourth point when not adjacent to the second or the fourth point of the image, and image-cutting a line passing through the selected valid second and fourth points. An image processing method characterized by using lines. 21. The method according to claim 20, further comprising an input step of inputting a band-shaped region position including a desired image cutout line and setting a center line of the band-shaped region as a predetermined line for the image. Image processing method. 22. In the line passing through the selected valid second and fourth points, there is no pixel having a density gradient equal to or higher than the predetermined threshold value in the second selecting step, and the fourth point is The method further comprises a line interpolation step of performing line interpolation on an unselected portion, wherein the line interpolation step is performed on both sides of a fourth point or a second point with respect to a portion where the fourth point is not selected. From the two nearest points, and a line segment between the intersections of two lines that pass through each of the two points and are substantially orthogonal to the center line and a line that surrounds the band-shaped region is defined between the two points. The image processing method according to claim 21, further comprising: 23. The smoothing step of applying a predetermined smoothing filter to the periphery passing through the selected second point and fourth point, and further comprising a smoothing step. The image processing method described in. 24. The image processing method according to claim 23, further comprising a step of displaying the image smoothed in the smoothing step. 25. The image processing method according to claim 23, wherein the smoothing step further strongly smoothes the periphery of the line interpolated in the interpolation step. 26. The method according to claim 14, further comprising a deforming step of deforming the image cutout line in a direction of the force by applying a predetermined virtual force to a predetermined position of the image cutout line. 25. The image processing method according to any one of 25. 27. A computer program product comprising a computer usable medium having computer readable program code means for cutting out a predetermined area from an image, said computer program product comprising: a first line on a predetermined line for the image. A computer-readable first program for selecting a second point that is closest to the predetermined point on a line that passes through the first point and is approximately orthogonal to the predetermined line and that has a density gradient of a predetermined threshold value or more. Code means and a pixel having a density gradient of a predetermined threshold value or more on a line passing through a third point on the predetermined line in the vicinity of the first point and orthogonal to the predetermined line A computer readable second program code means for selecting a fourth point closest to the selected second point, The second point and the fourth point selected in the first and second selection steps
A computer program product, characterized in that a line passing through the points is used as an image cutting line. 28. Further comprising third computer program code means for inputting a band-shaped region position including a desired image cutout line and setting a center line of the band-shaped region as a predetermined line for the image. A computer program product according to claim 27. 29. In the line passing through the selected second point and fourth point, the second computer program code means finds no pixel having a density gradient equal to or higher than the predetermined threshold value, The present invention further comprises fourth computer program code means for performing line interpolation of a portion where no point is selected, and the fourth computer program code means includes the fourth computer program code means.
2 points are selected from the 4th point and the 2nd point in the opposite directions with respect to the part where the point is not selected, and the two points pass through each of the 2 points and are substantially orthogonal to the center line. 29. The computer program product according to claim 28, wherein a line segment between the intersections of the line of the book and the line surrounding the band-shaped area is interpolated between the two points. 30. The method according to claim 27, further comprising fifth computer program code means for causing a predetermined smoothing filter to act on a periphery passing through the selected second point and fourth point. A computer program product according to any of paragraphs 29. 31. The computer program product according to claim 30, further comprising the step of displaying the image smoothed by the fifth computer program code means. 32. The computer program product according to claim 30, wherein the smoothing step further strongly smoothes the circumference of the line interpolated in the interpolation step. 33. A second pixel, which is on a line which passes through a first point on a predetermined line with respect to an image and is substantially orthogonal to the predetermined line, and which is closest to the point among pixels having a density gradient of a predetermined threshold value or more. A first computer program code means for selecting a point, and a predetermined threshold value on a line passing through a third point on the predetermined line in the vicinity of the first point and orthogonal to the predetermined line. Second computer program code means for selecting a fourth point closest to the selected second point among pixels having the above density gradient; and a predetermined number or more of the fourth points already selected. Third computer program code means for disabling the fourth point when not adjacent to another second or fourth point being stored, the selected valid second and fourth points The line that passes through the points of A computer program product characterized by: 34. The method according to claim 33, further comprising an input step of inputting a band-shaped region position including a desired image cutting line and setting a center line of the band-shaped region as a predetermined line for the image. Computer program products. 35. In the line passing through the selected valid second and fourth points, there is no pixel in the second computer program code means having a density gradient greater than or equal to the predetermined threshold, and the fourth The present invention further comprises third computer program code means for performing line interpolation of a portion where no point is selected, and the third computer program code means includes a fourth computer program code means.
2 points are selected from the 4th point and the 2nd point in the opposite directions with respect to the part where the point is not selected, and the two points pass through each of the 2 points and are substantially orthogonal to the center line. 35. The computer program product according to claim 34, wherein a line segment between intersections of a line of a book and a line surrounding the band-shaped area is interpolated between the two points. 36. The method according to claim 27, further comprising fourth computer program code means for causing a predetermined smoothing filter to act on a periphery passing through the selected second point and fourth point. The computer program product of any of paragraphs 35. 37. The computer program product of claim 36, further comprising the step of displaying the image smoothed by the fourth computer program code means. 38. The computer program product of claim 36, wherein the fourth computer program code means more strongly smooths the perimeter of the line interpolated by the third computer program code means. . 39. A fifth computer program code means for deforming the image cutout line in the force direction by applying a predetermined virtual force to a predetermined position of the image cutout line. A computer program product according to any of claims 27 to 38.