JP3673478B2 - Handwritten image output program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a handwritten tone image output program. More specifically, the present invention relates to a program for automatically converting and outputting an original image drawn in a mechanical tone by a drawing tool or CAD to an image drawn in a handwritten tone using a computer.
[0002]
[Prior art]
Conventionally, home dealers and interior goods dealers have drawn the layout of homes for sale and the layout of the sales interior goods using drafting tools or CAD, and distributed them as flyers or placed them in catalogs as books. It was. By using a drawing tool, CAD, or the like, each line has no distortion or protrusion, and the color has no spots (unevenness). In addition, for example, walls that are particularly emphasized in outlines are painted in solid black. A so-called mechanical image drawing in which lines and colors are drawn in a unified manner in this way is clear and easy to understand, and facilitates the creation of similar drawings.
[0003]
[Problems to be solved by the invention]
However, recently, drawings drawn by humans, that is, drawings with a little line distortion or color spots, are softer and more familiar than the mechanical drawings as described above. It has become popular because it is interesting and easy to realize. However, it would be time consuming and expensive for humans to write by hand.
The present invention has been made in view of the above-described problems, and an object of the present invention is to provide a program capable of automatically converting an image drawn in a mechanical tone by a drawing tool or CAD into a handwritten tone image.
[0004]
[Means for Solving the Problems]
In order to solve the above-mentioned problems, the invention of claim 1 is to convert an original image G0 drawn in a mechanical tone into an image G8 drawn in a handwritten tone, as well shown in FIGS. The handwritten-tone image output program TP of the first step S22, 23 for extracting the contour line r and the corner portion e of the contour line r in the original image G0 and the first step S22, 23 A second step S24 for projecting the outline r from the corner e, a third step S25 for extracting an emphasized outline rW to be emphasized from the outline r , and the emphasized outline extracted in the third step S25 A fourth step S26 for thickening the line rW ;
The image obtained in the second step S24 or the fourth step S26 is divided into a plurality of square child regions, and by applying a random number calculation to the vertex coordinates of each divided child region, each vertex position is shifted to each And a fifth step S4 for deforming the child region and distorting the contour r so as to correspond to the deformation of the child region .
According to the first aspect of the present invention, it is possible to express a state that is often common in handwriting, in which a straight line protrudes slightly at the corner e. Further, since the emphasized outline rW is thickened, the image is interesting.
Further, since the contour line r is deviated with respect to the edge of the base paper, for example, it is possible to express a state in which drawing is not performed well even if a ruler is used.
Further, since the plurality of line segments r1, r2,... In the contour line r are continuously declined, it is possible to express a state of drawing with freehand without using a ruler.
[0005]
In the second aspect of the invention, the second step S24 projects the contour line r at random from the corner e.
According to the invention of claim 2, since the contour line r protrudes randomly from the corner e, the feeling of handwriting is further enhanced.
The invention of claim 3 is a handwritten tone image output program TP for converting and outputting an original image G0 drawn in a mechanical tone to an image G8 drawn in a handwritten tone, as well shown in FIGS. The sixth step S1 for mixing the color in the original image G0 and the background color of the original image G0, the image G6 obtained in the first or second aspect, and the image obtained in the sixth step S1. And a seventh step S3 for synthesizing G2.
According to the invention of claim 3, in addition to the effect obtained by the invention of claim 1 or 2, the color in the original image G0 and the background color of the original image G0 are mixed, so that a color spot is formed. A state of drawing in a tone or colored pencil tone can be expressed.
[0007]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, a conversion program TP according to an embodiment of the present invention will be described with reference to the accompanying drawings.
The conversion program TP converts image data of a mechanical image input as raster data by an image reading device such as a scanner into image data of a handwritten image and outputs the image data. The conversion program TP is stored in a storage medium such as a floppy disk or a CD-ROM. Or you may make it download to the hard disk in a personal computer etc. from a suitable program delivery server.
[0008]
FIG. 1 is a diagram showing a comparison between an input original image G0 and a handwritten tone image G8 converted and output by the conversion program TP. FIG. 1A shows an image before conversion, and FIG. 1B shows an image after conversion.
As shown in FIG. 1A, the original image G0 is a mechanical tone image drawn in a mechanical tone using a drawing tool or CAD. That is, of the straight lines that are the contour lines, the horizontal line is parallel to the X axis along the vertical edge of the base paper, and the vertical line is parallel to the Y axis along the horizontal edge of the base paper. There is no distortion or protrusion in each line, the curvature of the arc is constant, and there are no spots on the painted part. It should be noted that the wall W, which is the part that particularly emphasizes the outline, is solidly painted with black. Other portions such as the floor and the desk T are drawn in brown or yellow.
On the other hand, in the handwritten tone image G8, as shown in FIG. 1B, the straight line that becomes the outline is distorted, and the outline protrudes at random from the corner e of each outline. In addition, there are spots on the painted parts, and a handwritten feeling appears. In the figure, sparsely drawn diagonal lines indicate that the color is uneven.
[0009]
Next, processing performed by the conversion program TP will be described with reference to FIGS. In order to make the description easy to understand, the description will be made by paying attention to the wall W and the desk T in FIG.
2 is a diagram schematically showing an image at each processing stage until the original image G0 is converted into the handwritten tone image G8 by the conversion program TP, and FIG. 3 is a flowchart showing an outline of processing contents performed by the conversion program TP. 4 is a flowchart showing the processing contents of the color conversion step S1, FIG. 5 is a flowchart showing the processing contents of the protrusion and enhancement step S2, FIG. 6 is a flowchart showing the processing contents of the distortion step S4, and FIG. 7 shows the L-shaped pattern LP. FIGS. 8A and 8B are diagrams for schematically explaining the processing performed in the distortion step S4.
[0010]
As shown in FIG. 3, the conversion program TP includes a color conversion step S1, a protrusion and enhancement step S2, an image composition step S3, and a distortion step S4.
In the color conversion step S1, the color of the wall W and the desk T in the original image G0 is converted into a color of watercolor or color pencil, and a color conversion image G2 is generated.
As shown in FIG. 4, the color conversion step S1 includes a spot generation step S11 and a blurring (blurring) step S12. In the spot generation step S11, the spot image G1 is generated by mixing the color of the wall W or the desk T in the original image G0 and the background color (for example, white). In the case of a watercolor painting tone, the color of the wall W or desk T and the background color are mixed at a mixing ratio based on predetermined parameters to adjust the saturation. In the case of a color pencil drawing tone, the color of the wall W or desk T and the background color are mixed at a random ratio for each pixel at a mixing ratio using random numbers. Since the spot image G1 looks rough as it is, it is made smooth by performing the process of the next tanning step S12. In the tanning step S12, the moving image method is applied to the spot image G1, thereby giving a check to the spot image G1. Thereby, the rough spot image G1 becomes a smooth color conversion image G2.
[0011]
In the projecting and emphasizing step S2, each contour r (rW, rT) of the wall W and the desk T and the corner e (eW, eT) of each contour r are extracted, and each corner e is random with a probability. An outline emphasis image G6 is generated by giving a protruding line pe (peW, peT) and thickening the outline rW on the wall W, which is the outline to be emphasized.
As shown in FIG. 5, the protrusion and enhancement step S2 includes a gray scale conversion step S21, an outline extraction step S22, a pattern matching step S23, a protrusion step S24, an enhancement part extraction step S25, and an enhancement step S26.
[0012]
In the gray scale step S21, the original image G0 is converted into a gray scale to generate a gray scale image G3. In the contour line extracting step S22, the contour line r of the gray scale image G3 is extracted, and a contour line image G4 drawn only by the contour line r is generated. The contour line r is extracted by a secondary differential operator. That is, for example, extraction is performed based on the difference in density of image data between adjacent lines in image data that is sequentially read in line units in the X or Y direction.
In the pattern matching step S23, a corner e in the contour image G4 is detected by pattern matching between the contour image G4 and L-shaped patterns LP1 to LP4 (see FIG. 7) created in advance as matching data.
[0013]
In the protrusion step S24, a protrusion line pe obtained by extending the contour line r forming the corner part e by a minute length is given to the corner part e detected in the pattern matching step S23 with a random probability. An outline protruding image G5 is generated. With the protruding line pe, it is possible to express a state often found in handwriting that a straight line has slightly protruded at the corner e.
In the enhancement portion extraction step S25, the portion to be enhanced, that is, the contour line of the wall W, is extracted based on the grayscale image G3 and the contour image G4. In the original image G0, the wall W is drawn dark in black, so that the portion is also dark in the grayscale image G3. First, by comparing the gray scale image G3 with a predetermined density threshold, the presence of the wall W on the image is known, and its position is specified. Next, in the contour image G4, a contour rW that overlaps the wall W is extracted as a contour to be emphasized. On the other hand, since the density of the desk T is low, the presence of the desk T on the image is unknown. Therefore, the outline rT of the desk T is not extracted as an outline to be emphasized.
In the emphasis step S26, the contour line emphasis image G6 is generated by thickening the contour line rW to be emphasized extracted in the emphasis part extraction step S25. Thereby, the wall W is expressed in a form surrounded by a thick outline, and the image is interesting.
[0014]
In the image composition step S3, the color conversion image G2 generated in the color conversion step S1 and the outline enhanced image G6 generated in the enhancement step S26 are combined to generate a composite image G7. The composite image G7 is expressed in a form in which the original image G0 is converted into a watercolor or color pencil tone, a random protrusion line pe is added to the corner e of the contour r, and the wall W is surrounded by a thick contour. It is an image.
By the way, at the time of composition, the contour line r becomes blurred due to the color conversion image G2. Therefore, in the synthesized image, the outline r and the outside thereof are once painted with the background color, and the outline-enhanced image G6 is overlaid again on the image, so that the outline r is not blurred.
[0015]
In the distortion step S4, a handwritten tone image G8 is generated by distorting the outline r of the composite image G7. Distortion is created by dividing the contour line r into a plurality of line segments r1, r2,... And declining the line segments r1, r2,.
As shown in FIG. 6, the distortion step S4 includes a coordinate data generation step S41 and a declination step S42.
In the coordinate data generation step S41, first, as shown in FIG. 8A, the composite image G7 is divided into a plurality of square child regions u, and the coordinates of the vertices of each child region u are obtained. In FIG. 8, the number of child regions u is set to 9 for the sake of easy understanding. However, the number is actually about 100. Next, by applying random number calculation to the coordinates of each vertex, the vertex of each child region u is shifted in the X direction and the Y direction. The random number calculation is performed, for example, by adding or subtracting a random number of a predetermined scale to the coordinate value of each vertex. By the random number calculation, each child region u is transformed into a distorted square as shown in FIG. In the declination step S42, the contour line segments rW1, rW2,..., RT1, rT2,. ,..., RT1 ′, rT2 ′,. The contour line r is deviated for each of the line segments r1, r2,..., RT1 ′, rT2 ′,..., Thereby expressing a state of drawing freehand without using a ruler.
[0016]
Further, when the contour line thickened in the emphasis step 26 is more than a predetermined amount in the child region u, the degree of deformation in the child region u can be increased. As a result, the distortion of the thick line portion is increased, and it can be expected that the image becomes more interesting.
Further, the anti-aliasing effect can be obtained by performing linear interpolation when deforming the contour line r in the child region u.
As described above, the description has been made with attention paid to the wall W and the desk T, but the same applies to the floor and the chair.
[0017]
As described above, according to the conversion program TP, the original image G0 drawn in mechanical tone, which is the input image, is colored in watercolor or color pencil, and the protruding line pe is randomly drawn at the corner e of the contour r. The handwritten tone image G8 is output by thickening the contour line rW of the wall W to be given with high probability, and further distorting the contour line r.
In the present embodiment, the processing method, processing order, processing timing, etc. in the whole or a part of the conversion program TP can be variously changed in accordance with the gist of the present invention.
[0018]
【The invention's effect】
According to the present invention, a mechanical image drawn by a drawing tool or CAD can be automatically converted to a handwritten image. Therefore, it is possible to obtain a handwritten tone image that is soft and familiar to the user and that is interesting and easy to realize without much time and cost.
[Brief description of the drawings]
FIG. 1 is a diagram showing a comparison between an input original image and a handwritten tone image converted and output by a conversion program.
FIG. 2 is a diagram schematically showing an image in each processing stage until an original image is converted into a handwritten tone image by a conversion program.
FIG. 3 is a flowchart showing an outline of processing performed by a conversion program.
FIG. 4 is a flowchart showing processing contents of a color conversion step.
FIG. 5 is a flowchart showing processing contents of a protrusion and emphasis step.
FIG. 6 is a flowchart showing processing details of a distortion step.
FIG. 7 is a diagram showing an L-shaped pattern.
FIG. 8 is a diagram for schematically explaining processing performed in a distortion step.
[Explanation of symbols]
TP conversion program (handwritten image output program)
r contour line r1, r2 multiple line segments e corner S1 color conversion step ( sixth step)
S22 Outline extraction step (first step)
S23 Pattern matching step (first step)
S24 Protrusion step (second step)
S25 Emphasis part extraction step (third step)
S26 Emphasis step (4th step)
S3 Image composition step ( seventh step)
S4 Distortion step ( fifth step)
rW Emphasized contour G0 Original image G8 Handwritten image (image)
G6 Outline-enhanced image (image)

Claims (3)

A handwritten image output program for converting and outputting an original image drawn in mechanical style into an image drawn in handwritten style,
A first step of extracting a contour line and a corner of the contour line in the original image;
A second step of projecting the contour line from the corner extracted in the first step;
A third step of extracting an emphasized outline to be emphasized among the outlines;
A fourth step of thickening the emphasized contour line extracted in the third step;
The image obtained in the second step or the fourth step is divided into a plurality of square child regions, and each child region is shifted by performing a random number operation on the vertex coordinates of each divided child region. And deforming the contour line so as to correspond to the deformation of the child region,
A handwritten tone image output program characterized by comprising:   The handwritten tone image output program according to claim 1, wherein the second step projects the contour line at random from the corner. A handwritten image output program for converting and outputting an original image drawn in mechanical style into an image drawn in handwritten style,
A sixth step of mixing the color in the original image and the background color of the original image;
A seventh step of synthesizing the image obtained in claim 1 or 2 and the image obtained in the sixth step;
A handwritten tone image output program characterized by comprising:

Images