JP2775122B2 - Automatic contour extraction vectorization processing method of illustration data and processing device used for the method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic contour extraction vectorization processing method for illustration data, and a processing apparatus used for the method.
Its by performing vector <br/> Torr process for the contour line, is compressed as much as possible the amount of information of the data necessary for illustrations of such illustrations and illustration extracts sharp contour of the boundary, In addition, the present invention relates to an automatic contour extraction vectorization processing method capable of quickly creating an illustration with a clear outline and easy to see, and an automatic contour extraction vectorization processing apparatus capable of quickly and accurately creating such an illustration.

[0002]

2. Description of the Related Art Recently, the use of computers has become remarkably diversified, and various results have been achieved in the field of image processing. Has a disadvantage that the processing time required to obtain a desired image (for example, an illustration) becomes very long.

In order to shorten the processing time required for creating illustrations, only information necessary for illustrations is extracted from image data obtained from original images such as photographs and video images, and other redundant information is removed. It is a general method to perform data compression by performing extraction processing to vectorize image data. Such feature extraction processing includes contour extraction processing, distance conversion / skeleton processing, thinning processing, intersection / end point detection processing, straight line / curve approximation processing, and the like. By performing these processes, data can be compressed, and the processing time required to create an illustration from an original image can be reduced.

[0004] However, if a large number of feature extraction processes as described above are performed, large data compression can certainly be achieved, but on the contrary, the shape features of the original image cannot be accurately restored.
The drawback is that the illustration becomes unclear . Therefore, if the feature extraction process is applied only to a part, the shape feature of the original image can be restored with high accuracy, but sufficient data compression cannot be performed this time. In other words, simultaneously satisfying both the reduction of the processing time for illustration creation and the improvement of the original image resilience acts as a trade-off problem, and it has been difficult to solve both problems in a well-balanced manner.

[0005]

SUMMARY OF THE INVENTION The present invention has been made in view of the above two trade-offs in the creation of illustrations, and is intended to form a clear outline of a boundary from an original image such as a photograph or a video image. by performing <br/> vector processing for extracting and its contour, is compressed as much as possible the amount of information of the data necessary for illustrations of such illustrations and illustration, moreover, the contour is sharp easy to see Illustration It is an object of the present invention to provide a method for automatic contour extraction and vectorization processing of illustration data, which can quickly generate an image.

Another object of the present invention is to provide an automatic contour extraction vectorization processing apparatus for illustration data capable of quickly and accurately creating such illustrations.

[0007]

In the present invention SUMMARY OF THE INVENTION, generates image data are hierarchically classified by the difference in brightness or color from the original image, the brightness or color from the image data above a predetermined threshold <br/> value The points where different hierarchical regions are in contact are extracted to generate provisional image contour information.
Select an indistinct outline of the information and select the area
And change the threshold of this area to
Boundary points are extracted by watermarking and re-extracting
After generating the sharp image contour information, thus generating
By performing distance conversion and skeletal processing on the figure in the obtained image outline information to automatically generate the line width information of the outline included in the original image, the figure in the image outline information is thinned. Generate a center line, generate a connection line closer to the contour shape of the original image by re-extracting the connection line by comparing the line width and the line width information of the connection line once extracted from the center line, By approximating the line drawing figure composed of the connecting lines with a curved line and a straight line represented by mathematical formulas, data-compressed vector contour information is automatically generated, and by adding the line width information to the vector contour information, an original image is obtained. A processing method for solving the above-mentioned problem is provided by adopting a means of reconstructing essential contour information components included as illustration data.

[0008] In the present invention, generates layer classified image data to the difference in brightness or color from the original image, the bright <br/> dark or color tone from the image data equal to or larger than a predetermined threshold value is different hierarchies region Extract the contact points
Generate temporary image outline information, and
Select an unclear outline partially to specify its area,
Contour that hides in the unclear area by changing the threshold of the area
Means for generating image outline information with clear boundaries by watermarking and re-extracting points ; distance conversion and skeleton processing for the figure in the image outline information thus generated ; Means for automatically generating line width information of a contour line included in an original image; means for thinning a figure in the image contour information to generate a center line; Means for comparing the line width with the line width information and re-extracting the connection line to generate a connection line that is closer to the contour shape of the original image; Means for automatically generating data-compressed vector contour information by approximation with the following; by adding the line width information to the vector contour information, essential contour information components contained in the original image are illustrated. By adopting a means of reconfiguring as a data, a processing apparatus that solves the above problem is provided.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in more detail based on embodiments shown in the accompanying drawings. FIG. 1 is a schematic block diagram of an apparatus according to an embodiment of the present invention, FIG. 2 is a flowchart showing an automatic contour extraction vectorization processing method of the present invention, and FIG. flowchart showing all 11 steps of the curve separation process, FIG. 4 is 4- vicinity of illustration, FIG. 5 8 near the illustration, FIG. 6 is an explanatory view of the distance b _d.

First, as shown in FIG.
(Manufactured by NEC; trade name PC-9801DA), and the display device 2 (manufactured by NEC; trade name PC-KD88).
2), Laser printer 3 (manufactured by NEC; trade name PC)
-PR602PS), the image scanner 4 and the video image processor 5 are connected. Further, a video camera 6 and a video monitor 7 are connected to the video image processor 5. Among them, the computer 1 incorporates a dedicated substrate developed by the present applicant, and enhances image processing capability. The video image processor 5 is a device for converting a video image signal into a signal that can be handled by the computer 1.

Using this apparatus, an original image, for example, a photograph, is read several times by the image scanner 4 and data sent from the image scanner 4 as a set of points indicated by the hierarchical values is input to the computer 1. By averaging the hierarchical values of the points, predetermined hierarchical image data is obtained. Treat this image data as new input data and exceed the predetermined threshold
Will be stored as an image contour information provisional while extracting the different parts of the hierarchy region is in contact with shading to. After this processing, the image temporary outline information is displayed on the display device 2 and it is checked whether or not the image temporary outline information includes extra outline information. By using the mouse shown in FIG. 1 and further increasing the threshold value at the time of extracting the image temporary contour information, extra contour lines are removed. Also, if a partially unclear outline appears, the threshold value is changed by limiting the area including the part with the mouse , and the outline points hidden in the unclear area are extracted again.
It may do it with. Then, the unknown in the provisional contour information
Clear data elements are clarified and overall boundaries are sharp
Data is generated, and the generated outline is a line drawing figure displayed in binary, which is used as image outline information.

By performing the contour extraction process on the hierarchical image data of the multi-valued image obtained from the original image,
The amount of data required for illustrations such as illustrations and illustrations can be significantly reduced compared to image data.In addition, when creating illustrations, it is better to use outline data extracted from image data showing the entire original image On the contrary, a clear and easy-to-read illustration can be obtained. Further, since the data is compressed to a greater extent than the image data, the processing time is short even when the illustration created once is modified / changed, and the modification / change is easier than the illustration composed of the image data.

When it is desired to draw an illustration directly from a prototype or the like, the prototype is directly imaged by the video camera 6 while sufficiently illuminating from a plurality of directions so as not to generate extra shadows. After the image signal is converted into a signal that can be handled by the computer 1 by the video image processor 5, the signal converted by the image processor 5 is regarded as data consisting of a set of points represented by hierarchical values, and Performs the same contour extraction processing as
What is necessary is just to generate image outline information. In this way, when drawing illustrations directly from the actual thing, using a video camera saves time and effort such as developing a photograph.

Thereafter, in order to further compress the data, vectorization is performed on the image outline information with a clear boundary generated as described above. As shown in FIG. 2, the vectorization processing in the present embodiment includes 1) a distance conversion / skeleton processing step, 2) a thinning step, 3) a connection line extraction step, 4) a connection line reprocessing step, and 5) a straight line. It comprises a total of eight steps: a curve separation step, 6) a connecting line integration step, 7) an equation approximation step, and 8) a vector contour information generation step. Of these, only the distance conversion / skeleton processing is not the processing for the purpose of data compression, but is the processing performed to generate line width information in the outline included in the original image as a personality factor characterizing the illustration. The reason for performing such line width information generation processing is that if only the vectorization processing required for data compression is performed, the shape characteristics of the original image are almost lost, and only a simple shape illustration with little accent is obtained. Because.

1) Distance Conversion / Skeleton Processing Step In this step, distance conversion / skeleton processing is performed on the given image outline information to automatically generate line width information. Although the normal purpose of the distance conversion and skeleton processing is to restore the original figure from the obtained skeleton and distance values, the distance conversion and skeleton processing of the present embodiment uses the contour lines included in the original image. The purpose of the present invention is to generate line width information in, and to use this line width information as an individuality factor characterizing the illustration to draw an illustration composed of easily-viewable contours with perspective and accent. Distance conversion of this embodiment
The skeleton processing will be described as follows. First, for a figure in the image outline information composed of white and black grids, 0-pixels (= white point) and 1-pixels (= black point), from the upper left to the lower right (forward direction) and from the lower right Raster scanning is performed twice to the upper left (reverse direction) to perform distance conversion processing. This distance conversion is a process for obtaining the shortest distance to the 0-pixel in each pixel of the figure, and the value of each pixel is converted so as to take a higher value toward the center of the figure. Next, in the distance-converted image, a set of pixels having a maximum distance value is defined as a skeleton. This skeleton is a linear set located at the center of the figure, and this line substantially matches the center line obtained by the thinning processing described later. Then, the 1-pixel point adjacent to the other 0-pixel and 1-pixel points is removed, that is, the pixel value of the 1-pixel point is replaced with 0, and the pixels removed until reaching the skeleton are removed. The number, that is, the number of pixel values replaced with 0 is integrated, and this is determined as the line width. (Note that this line width information is temporarily stored until it is used in the connection line reprocessing process described later.)

2) Thinning process In this process, the line width of the outline is reduced by repeatedly removing 1-pixel points satisfying a specific condition from the figure in the given image outline information. This is a process of obtaining a center line having a line width of 1. As a method of obtaining this center line, a skeleton obtained by the above-described distance conversion / skeleton processing can be used as the center line.However, since this skeleton has a drawback that connectivity of the original figure is not completely preserved, In this embodiment, thinning processing is employed. In the thinning processing of the present embodiment, Deutsch, one of conventionally known processing methods, is used.
(Refer to “Introduction to Computer Image Processing”, supervised by Soken Publishing, “Hideyuki Tamura”, pp. 80-83), and as shown in FIGS. 4 and 5, one-pixel points to be removed must be satisfied. Basically, the condition is 4-neighbor (pixels next to vertical and horizontal)
Under one or more 0-pixels and one or more 1-pixels in the 4-neighbor, and the number of 1-pixels in the 8-neighbors (vertically, horizontally and diagonally adjacent pixels) is 2 or more is there. (A point that satisfies this condition is a point on the boundary between the black and white of the figure and not an end point of the line.) However, for a line having a width of 2, a 1-pixel point is obtained under the above condition. If all are erased, the line disappears, so a further special condition is imposed, so that only the 1-pixel points arranged on one side of the line are removed to obtain a line having a width of 1.

3) Connection Line Extraction Step The purpose of this step is to extract a continuous line from the center line obtained in the thinning step.

Actually, first, 3) 1) a certain 1-pixel point is extracted, and 3) 2) whether or not there is a 1-pixel point in the 8-neighborhood of the 1-pixel point is scanned. If there is a 1-pixel point in the 8-neighborhood, the process proceeds to 3), and if there is no 1-pixel point in the 8-neighborhood, the extraction of the connection line having the 1-pixel point as an end point ends. Go forward respectively. Thus, if 3-2) finds a 1-pixel point in the 8-neighborhood, 3-3) extracts one of the 1-pixel points, then 3-4) this newly extracted Scans whether there is a 1-pixel point other than the 1-pixel point previously extracted or a 1-pixel point crossing in the 8-neighborhood of the 1-pixel point, and if the 1-pixel point satisfies the condition If there is, go to 3-5), and if there is no 1-pixel point that satisfies the condition, the extraction of the connection line ends. When there is a 1-pixel point that satisfies the condition in 3-4), 3-5) one of the 1-pixel points is extracted and the point is set as a scanning target of a neighboring pixel point.
It returns to 3-4) again. Note that the intersection 1-pixel point refers to an intersection such as when the connecting lines cross each other.

By repeating the operations of 3-4) and 3-5), it is possible to extract a connection line having two or less end points in a linear or annular shape or a slightly more complicated shape (for example, a figure of 6). is there.

However, only one connection line can be extracted by simply repeating the operations of 3-4) and 3-5). Therefore, after the extraction of one connection line is completed, that is, 3-6) 1-pixel points that have not yet been extracted in the connection line extraction work so far are extracted, and from this 1-pixel point, The repetition of the operations of 3-4) and 3-5) is started. Then, when it is finally determined that all the 1-pixel points have been extracted by any one of the connection line extraction operations and there are no remaining 1-pixel points that have not been extracted any more, the connection line extraction is ended.

4) Connection Line Reprocessing Step The connection line extracted in the connection line extraction step is recognized as one connection line divided into two or more connection lines or as another connection line. There is a possibility that two or more lines to be performed are recognized as one connection line. Therefore, information on the thickness of the contour line (line width information) obtained in the distance conversion / skeleton processing process is given to the extracted connection line, and 1
When the thickness suddenly changes at a certain point of the connection line, the connection line is divided into two at the point. Next, among the connecting lines having the thickness information obtained in this way,
Take out one of the same thickness, and if there is an end point in the connection line that is adjacent to the other connection line end point, connect this, or the end point in the connection line sandwiches the other 1-pixel point If it is adjacent to the end point of another connection line, it is connected. This makes it possible to generate a connection line closer to the contour shape of the original image. As described above, in the present embodiment, the line width information generated in the distance conversion / skeleton processing process is not only used as an individuality factor characterizing the illustration,
It is also used to extract connection lines with higher accuracy.

5) Straight Line / Curve Separation Process The straight line / curve separation process is basically a process of repeating a work of separating a curve having a certain curvature or more from a connecting line while changing a threshold value.

In describing this process, some introduction is required. Pixel distance d _k ; Number of pixels sandwiched between two pixels on connection line D plus 1 Sequence {d _k }; Sequence with pixel distance d _k as the k-th term Pixel point P _i ; Connection line On D, the i-th pixel point from the scanning end point Pixel point P _{i + d} ; On the connection line D, the i + d-th pixel point from the scanning end point Pixel point P _id ; When i−d ≧ 1, the connection line on D, when the scanning end point i-d-th pixel point i-d <a 1, P ₁ and the pixel point spaced advanced by the virtual pixel number d from P _i on the straight line connecting the P _i to P ₁ Distance b _d ; distance between pixel point P _i and a straight line connecting pixel point P _{i + d} and pixel point P _id with respect to a certain pixel point P _i on connection line D Threshold B; constant Number of constituent pixels n; connection Total number of pixels of line D

Now, this process will be briefly described. 5-1) Condition setting, 5-2) Calculation of separation distance b _k , 5-3)
Separation of curves and straight lines, 5-4) Integration of curve components and linear components, 5-5) Re-evaluation of curve components, 5-6) Arrangement of curve components, 5-7) Division of curve components, 5-8) It consists of a combination of all 11 procedures of re-evaluation of linear components, 5-9) integration of linear components and curve components, 5-10) determination of re-separated components, 5-11) resetting of conditions.

5-1) Setting of Conditions Before the separation of a curve and a straight line, a connection line D to be processed and an initial value of a variable k are set. First, one of the connecting lines determined to have the same line width in the connecting line reprocessing process is taken out, and then k = 1 is set, and the following 5-2) is performed.

5-2) Calculation of separation distance b _k A plurality of appropriate pixel points P _i on the connection line D are extracted, and d = d _k
Is calculated as the separation distance b _d . However, the extracted pixel points are separated by a distance 2d.

5-3) Separation of Curve and Straight Line Each of the separation distances b _d obtained in step 5-2) is compared with a threshold value B. If b _d ≧ B, [P _id P _{i + d} ] Is separated into a curve component D ^C , and to 5-4) If b _d <B, [P _id P _{i + d} ] is separated into a linear component D ^{L and then} to 5-8) (where [P _id P _{i + d} ] is on the connecting line D,
(Closed section from pixel point P _id to pixel point P _{i + d} )

5-4) Integration of Curve Components and Linear Components In this procedure, components determined to be dependent components in step 5-8) described later are connected to adjacent curve components.

[0029] 5 5) to the curve component D ^C obtained through the integration of 4) curve component and a linear component of the curve reevaluation of components 5, reevaluated performed. That is, both ends of the curve component D ^C, together determine whether a straight line component is present, when either parameter which linear component of the two ends of the curve component D ^C is not present, to determine the curve component D ^C a curve component, When the linear component exists at both ends of the curved component D ^C , when the linear component exists at both ends of the curved component D ^C , the number n of constituent pixels of D ^C ≧ d _k ∧2 component−3 component ratio R <
If the threshold value N, to determine the curve component D ^C a curve component, when other than the above to 6) of 5, is determined as a dependent component D ^C, 5 of 7) to (provided that 2-component -3 component ratio R Is the curve component D
The total number of pixels of the two linear components that are present at both ends and ^C, divided by the total number of pixels of the three components of both linear component and a curved component D ^C.
The threshold value N is a constant satisfying 0 <N <1. )

5-6) Arrangement of curves 5) 5) Those which have been determined to be curve components in the re-evaluation of the curve components are smoothly connected, and 5) straight line / curve separation at the connection line of the line width concerned. End all steps of the process. Then, returning to 5) 1) again, the 5) straight line / curve separation process is performed for other connection lines for which the curve / straight line separation is not performed.

[0031] In 5 of 7) 5 divided curve component 5) re-evaluation of the curve component, the component D ^C it is determined that the dependent component is divided into two, into 5 9).

5-8) Re-evaluation of linear component 5-3) Re-evaluation is performed on the linear component D ^L separated by curve / linear separation. That is, it is checked whether a curve component exists at one of both ends of the linear component D ^L , and the linear component D ^L is determined.
When the curve component is present next to ^L, the number of constituent pixels of D ^L n
If ≦ 2d _{k + 1,} it is determined that D ^L is a dependent component and 5-4)
To other than the above, go to 5-9)

5-9) Integration of linear component and curve component The component obtained by dividing the component into two in 5-7) Curve component is connected to adjacent linear components.

5-10) Determination of re-separated components From the linear components obtained through the integration of the linear components and the curved components, those which need re-separation are extracted. That is, D
^The number n of the constituent pixels of ^L is checked. If n ≦ 2d _{k + 1} , D ^L is determined to be a linear component. If n> 2d _{k + 1} , D ^L is determined to be a re-separable component and 5
To 11)

5-11) Condition resetting 5-10) From the determination of the re-separation component, one of the components determined to be the re-separation component is taken out, and this component is regarded as a new connection line and is set to D. Subsequently, k + 1 is replaced with k, and the process returns to 5) 2). The other re-separation required components are processed after the straight / curve separation of the new connection line D is completed. When 5-10) re-separation components are not determined in the determination of re-separation components, go to 5-6).

In step 5-3), the distance b
_d is assumed to be proportional to the curvature of [P _id P _{i + d} ],
It is compared with a threshold value B. Assuming that [P _id P _{i + d} ] forms a complete arc, [P _id P _{i + d} ] when b _d = B
The curvature C _k of given by the following equation.

[0037]

(Equation 1)

Further, in step 5 of 5), although D ^C constituent pixel number n and the pixel distance d _k of interest is being compared, when number of pixels of approximately D ^C is less than the pixel distance d _k is a D ^C Low value for separation as independent curve components.
In this case, it is preferable to connect both ends either component or both components D ^C. Therefore, when the linear component is present at both ends of the D ^C, after divided into two D ^C in 5 of 7), it is to be connected to the linear component 5 9).

Furthermore, in 5-5), two components-
The three-component ratio R and the threshold N are also compared.
It is intended to consider the value to separate D ^C as a curve component. If either the even linear components across the R ≧ N and D ^C is present, more linked to the linear component of both ends divided into two D ^C is a smooth curve can be easily obtained later.

In 5-8), the number n of pixels constituting D ^L is compared with 2d _{k + 1} . The reason is that, for a pixel that is once separated as a linear component (that is, b _k <B) and the number of constituent pixels is smaller than 2d _{k + 1} , the pixel distance is set to d
_{Even if k} is changed to d _{k + 1} , b _{k + 1} <B is always satisfied, and it is no longer worthwhile to perform 5-3) Separation of Curve / Line. Therefore,
D ^L is connected to an adjacent curve component or is determined as a straight line component.

In this embodiment, the sequence {d _k }
Is given by the following equation:

[0042]

(Equation 2)

6) Integration of connecting lines In this process, a bending point is detected from the preceding process, that is, 5) curve components separated by curve / straight line separation, or
Obtain longer curves and straight lines from the curve and straight line components.

First, for those classified as curve components,
In the pixel point in the curve component, of 5 of 2) points are calculated b ₁ when calculating the distance b _k, extracts the points most value of b ₁ is greater, the pixel point P _i , Two points P _{i−d ′} and P _{i + d ′} separated by the special scanning pixel distance d ′ given in advance are extracted. Then, calculate the inner product of as vectors follows, if the inner product is greater than the threshold value I, the pixel point P _i and bending points determined. At this time, the curve component is divided by the pixel point P _{i, and if} the curve component is adjacent to another curve component or linear component, the divided component is connected to the adjacent component.

[0045]

(Equation 3)

Thereafter, the curve components for which no bending point is detected are connected if they are adjacent to other curve components. Thus, a long curve component is obtained.

For the straight line components, two straight line components are taken out of all the straight line components, and the distance S between the closest ends of the straight line components and the distance D 'between the furthest both ends are calculated. After calculating the lengths L ₁ and L _{2 of} the components and the angles θ ₁ and θ ₂ formed by the straight line connecting both ends of the two straight lines and the two straight line components, these are introduced into the determination formula,
If it is larger than the threshold value, the two straight line components are connected at their nearest end points. If this process is repeated, it is assumed that the linear component is finally determined. Note that the determination formula is as follows.

[0048]

(Equation 4)

7) Formula approximation process In this process, the curve components and the linear components finally classified in the previous process are approximated by applying them to mathematical formulas. Linear components are approximated by the equation of a straight line connecting both ends of the linear components. Curve components are appropriately divided into several sections, and then both ends of each section and three-minute points are used for a total of four points. To a Bezier curve passing through.

The point on the approximated Bezier curve is
^{_{(A x t 3 + b x}} t 2 + c x t + d x, a y t 3 + b y t 2 + c y
t x d _y ), and a _x , b _x ,... indicate that the coordinates of the four points are (P _0x , P _0y ), (P _1x , P _1y ), (P
_2x , P _2y ) and (P _3x , P _3y ) are given by the following equations.

[0051]

(Equation 5)

8) Generation process of vector outline information As described above, 2) thinning process to 7) approximation process by mathematical formula
The sharp image outline information of the boundary generated by extracting the outline of the original image is automatically vectorized and further converted into data-compressed vector outline information. By adding the line width information generated in the above-mentioned 1) distance conversion / skeleton processing process to this vector outline information, only the essential outline information component from which the extra information included in the original image has been removed is re-used as illustration data. Can be configured. If vector outline information with such line width information is used, the amount of data required for the illustration is greatly reduced, so that processing time can be reduced and used as an individuality factor that characterizes the outline of the illustration. Since the line width information includes the contour line, the user can draw an illustration composed of easy-to-see contour lines with perspective and accent.

The present embodiment is as described above. However, when approximating the separated curve components by a mathematical expression, approximation to a curve other than a Bezier curve, for example, a spline curve, an arc curve, or a conic curve is performed. Can also.

[0054]

As it is evident from the foregoing description, in the present invention method,
Provisional image outlines from original images such as photos and video images
Generated and partially select the indistinct contour
To specify the area, change the threshold for this area,
The contour points hidden in the unclear area are watermarked and extracted again.
Having adopted skilled outline extraction process of that, illustrations and sharp image contours of the boundary suitable for illustrations of such illustration
Information is obtained. Furthermore, the thus obtained sharp boundaries
Vectorization processing such as distance conversion, skeleton processing , thinning, and mathematical expression approximation processing is performed on the contour information.
The operator changes the contour line during the turretization process to make it clearer.
Processing can be performed automatically without the need to reduce the processing time required to generate vector contour information.
In addition, this vector contour information is vectorized
Because the original image outline information targeted for is clear,
The illustration drawn with the vector outline information is clear
Easy to see. The contour extraction process that has been devised in this way
By combining with motion vectorization processing,
It became possible to quickly create clear and easy-to-read illustrations.
You.

In the vectorization processing, the line width information generated in the distance conversion / skeleton processing is compared with the line width of the connection line in the thinning / mathematical expression approximation processing to re-extract the connection line. Therefore, the contour shape of the original image can be restored with high accuracy. Furthermore, the line width information is
Since the contour information is added to the contour information , it is possible to use the line width information to create an illustration composed of an easy-to-see accented outline , thereby further improving the sharpness of the illustration. like this
Effective use of line width information for original image restoration
It became possible to further improve the sharpness above.

According to the apparatus of the present invention, pictures and videos
Generates image outline information with sharp boundaries from original images such as images
Form, automatically generate vector outline information and vectorized is data <br/> data compressing the contour information, the vector
Since the essential contour information component contained in the original at the contour information can be reconstructed as illustration data, outline crisp legibility
There it is possible to greatly improve the work efficiency in the creation of illustrations. As described above, the method and apparatus of the present invention have many effects in the field of illustration creation, and have a very large practical value.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of an apparatus according to an embodiment of the present invention.

FIG. 2 is a flowchart illustrating an automatic contour extraction vectorization processing method according to the present invention.

FIG. 3 is a flowchart showing all 11 procedures of a straight line / curve separation process in the automatic contour extraction vectorization processing method of the present invention.

FIG. 4 is an explanatory diagram of a 4-periphery.

FIG. 5 is an explanatory diagram of an 8-neighborhood.

6 is an explanatory view of the distance b _d.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Computer 2 Display device 3 Printer device 4 Image scanner 5 Video image processor 6 Video camera 7 Video monitor

 ────────────────────────────────────────────────── ─── Continuation of front page (56) References JP-A-2-100768 (JP, A) JP-A-62-173472 (JP, A) JP-A-60-132280 (JP, A)

Claims (2)

(57) [Claims] 1. A generates layer classified image data to the difference in brightness or color from the original image, the image contour of the temporary brightness or color from the image data above a predetermined threshold value to extract the differences hierarchical region are in contact information
Generate and partially remove unclear contours from this temporary contour information.
Select and specify the area, change the threshold for this area
Extract and re-extract the outline points hidden in the unclear area
After that, the image contour information with a clear boundary is generated, and then the figures in the image contour information thus generated are subjected to distance conversion and skeleton processing to automatically generate the line width information of the contour lines included in the original image. On the other hand, thinning a figure in the image outline information to generate a center line, comparing the line width of the connection line once extracted from the center line with the line width information, and re-extracting the connection line. A contour line closer to the contour shape of the original image is generated by using the above-described method, and a line drawing figure composed of the connection line is approximated by a curve and a straight line represented by mathematical formulas, thereby automatically generating data-compressed vector contour information. Automatically reconstructing essential contour information components contained in the original image as illustration data by adding the line width information to the original image. Vectorization processing method. Wherein generating a hierarchical classified image data to the difference in brightness or color from the original image, the image contour of the temporary brightness or color from the image data above a predetermined threshold value to extract the differences hierarchical region are in contact information
Generate and partially remove unclear contours from this temporary contour information.
Select and specify the area, change the threshold for this area
Extract and re-extract the outline points hidden in the unclear area
Means for generating image contour information with clear boundaries by performing distance conversion and skeletal processing on the figure in the image contour information thus generated to automatically generate line width information of contour lines included in the original image Means for thinning a figure in the image outline information to generate a center line; and comparing the line width of the connection line once extracted from the center line with the line width information to determine the connection line. Means for generating a connected line closer to the contour shape of the original image by re-extracting; automatically generating data-compressed vector outline information by approximating a line drawing figure formed by the connected line with a curve and a straight line represented by mathematical expressions Means for reconstructing essential contour information components contained in the original image as illustration data by adding the line width information to the vector contour information. Automatic contour extracting vector processing apparatus of illustration data to.