JPS61221979A - Graphic processing system - Google Patents

Abstract

PURPOSE:To compress information with a high degree of approximation by obtaining intersecting points in a thin line graphic form in particular as the primary features out of a graphic form handwritten freely on a form, etc. and converting the screen information on the graphic form into a vector. CONSTITUTION:In a graphic system where the graphic information is compressed for display, a line thinning process is applied first to an input graphic form by a line thinning means 1 for production of a thin line graphic form. Then a local intersecting point extracting means 2 extracts the local feature points to the thin line graphic form. A stroke tracing means 4 of an information compressing part 3 traces strokes from the local feature point to extract the local inflection points and local terminal points in relation to the corresponding stroke and also to decide the length of each line element. Then an overall feature extracting means 5 performs collation with the porescribed threshold value based on the information on the length of the element line to disuse the undesired local inflection points and local terminal points and decides the overall intersecting points and overall terminal points for compression of information.

Description

[Detailed Description of the Invention] [Summary] The technical outline of the present invention is as follows. It involves a system that compresses and expresses the graphic information of a drawn figure based on local feature points.The system traces the stroke starting from the intersection of thin line figures and calculates the local inflection points and line elements. The present invention relates to a graphic processing system that extracts distances, determines global inflection points from the obtained local inflection points, and compresses information to the extent that the characteristics of the drawn graphic are not impaired.

[Industrial application field]

The present invention mainly relates to a graphic processing system that uses a digital computer to read free hand drawn figures, extract features, and compress information.

[Prior art and problems to be solved by the invention]

Track strokes between local feature points (intersections, end points),
Graphic processing systems with improved processing for determining global inflection points have already been proposed (for example, Japanese Patent Application No. 53-15
No. 9549 "Graphic Processing System").

This graphic processing system is a graphic processing system that expresses a drawn figure by compressing the figure information of the figure based on the local feature points of the figure. Starting from point i, the stroke is traced, the distance between the local inflection point and the line element is extracted, and the global inflection point is determined from the obtained local inflection point. This is a graphics processing system that compresses information to the extent that the characteristics of the drawn graphics are not lost. In other words, the method for inputting a drawn figure into a data processing device is to input all the information regarding the figure to the data processing device once, and then perform figure processing, such as enlarging or reducing the figure globally or locally. When performing man-machine interactive operations such as rotation and addition of figures using a display etc., the above-mentioned drawn figure is converted into a form suitable for processing within the data processing device and inputted. This is a graphic processing system that compresses information mainly by utilizing the coordinates of intersection points of graphics, omitting the human-machine interactive operations as much as possible. However, in such a figure processing system, if the stroke between local feature points has a series of minute line elements that gradually curve significantly and line elements that are long to some extent, the approximation accuracy may be affected depending on the degree to which they are mixed. The problem was that it fell off.

This will be described in more detail with reference to FIG.

FIG. 2 is a diagram showing one stroke between local intersection points, which is the basic unit of approximation. In the figure, ■ marks indicate local intersection points, ○ marks (including * marks) local inflection points, and * marks extracted global inflection points. If we try to approximate this stroke using the extracted global inflection points, the local intersection p1
Although there is an inflection point between the global inflection points 02 and 02, it cannot be detected, and until then, the figure cannot be accurately approximated.

[Means and effects for solving the problem] The present invention has the following features:
In order to solve the above problem, by examining in detail the local inflection points between the global inflection points extracted based on the global inflection points extracted in the same way as before, The idea is to calculate the degree of difference between the figure and the figure, and if the degree is large, to generate a new global inflection point, which enables accurate information compression of the figure.

According to the present invention, in a figure processing system that compresses and expresses figure information based on local feature points of a drawn figure as intersection points, end points, and one point on each closed loop, thinning is performed to thin the figure. processing means, local feature point extraction means for extracting local feature points from the thin line figure obtained by the thinning processing unit, and local feature point extraction means for extracting local feature points from the thin line figure obtained by the local feature point extraction means. traces at least one stroke of the stroke and extracts a local inflection point on the stroke until it reaches another local feature point, and also extracts a local inflection point between the local inflection points or between the local feature point and the local inflection point. a stroke tracing processing means for extracting the distance of a line element between the points, and a global inflection point by comparing the distance with a threshold value and collecting the local inflection points that are adjacent to each other within the threshold value. extract,
A global feature extraction processing means is provided for generating a new global inflection point according to the degree of difference between the global inflection point and a thin line figure based on the global inflection point, and the stroke between the local feature points is A graphic processing system is provided, characterized in that the graphic processing system is represented by at least the feature point coordinates and the global inflection point coordinates.

〔Example〕

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

FIG. 1 (al indicates an embodiment of the configuration of a thinning processing section 1 forming a thinning processing means of the present invention. Reference numeral 1 in the figure
8. 19 are respective sub channels; 20 is an address control unit which generates an access address for the input video 21 according to instructions from the sub channel 18;
1 is an input video whose processed figure corresponds to image information; 22 is a module control unit that performs a thinning mask (filter) to be described later based on instructions from the subchannel 19;
23 is a thinning mask (filter) circuit section that operates a program module that puts the circuit section 23 into operation, and performs line thinning on the video data that is successively outputted as the input video 21. The reference numeral 24 indicates a thin line figure which is thinned by the circuit 23 and is held in correspondence with the address of the input video 21.

That is, the thin line figure shown in FIG. 2 is obtained by the thin line processing section 1.

FIG. 1 (bl indicates an example configuration of the processing section 2 constituting the local feature point extraction processing means. Reference numeral 18' in the figure
, 19' are subchannels, 20' is an address control section, 24 is a thin line figure corresponding to FIG. 26 is a local feature point extraction mask (filter) circuit that operates the program module that puts the extraction mask (filter) circuit unit 26 into operation, and also applies to thin line figure data that is sequentially read out as thin line figures. 2. Extracting local feature points by
Reference numeral 7 indicates a feature point video in which extracted local feature points are held in correspondence with the existing address of the thin line figure.

FIG. 1 (C1 shows an embodiment of the local feature point coordinate extraction section, which extracts the coordinates of each local feature point based on the feature point video 27 obtained by the configuration of mountain 1 in FIG. 1. Extract the coordinate position. Reference numbers 18'' and 19'' in the figure are subchannels, 20# is the address control section, 27 is the feature point/video corresponding to FIG. 1 ('b), and 28 is the module control section. 29 is a coordinate point extraction circuit section which operates a program module that activates a coordinate point extraction circuit section 29, which will be described later, in response to an instruction from a subchannel 19'. A system is shown in which when a feature point is read out, the local feature point is extracted and the coordinate information of the feature point is set in the feature point memory 30.

FIG. 1 (dl indicates the configuration of an embodiment of an information compression processing unit having a stroke tracking processing means and a global feature extraction processing means, symbols 1B"' and 19"' in the figure are subchannels, respectively, and 20 is an address. The control unit 24 is shown in FIG.
), 30 is a feature point memory corresponding to FIG. 1(C), 31 is a microprocessor, and 32 is a coordinate point memory for storing the coordinate points of the figure to be processed.

The graphics processing system is adapted to perform processing as shown in the form of a flowchart in FIG. That is, (1) Thinning is performed on the input figure (figure to be processed) in the thinning processing means 1 to create a thin line figure.

(2) The local intersection extraction processing means 2 extracts local feature points from the thin line figure using a feature point extraction mask.

(3) The stroke tracking processing means 4 in the information compression processing unit 3 performs full stroke tracking starting from 10 local feature points, extracts local inflection points and local end points related to the stroke, and extracts the length of each line element. determine the

(4) Next, in the global feature extraction processing means 5, (i
) Based on the length information of the line elements, sort out the unnecessary points among the local inflection points and local end points by comparing them with the threshold value, and (ii) also sort out the unnecessary ones among the local intersection points. (iii) determine global intersection points, global points of inflection, and global end points; and (iv) perform information compression.

(5) The above process (3) for one stroke.

When (4) is completed, a label is energized for the stroke, and the same process is repeated for all strokes between the local feature points.

(6) When the processing for all strokes between local feature points is completed, the label detection processing means 6 detects the presence or absence of strokes to which no labels are attached, and executes the above processing +31 (4). . Note that the above process (
At the stage where step 5) has been completed, the stroke to which no label has been assigned means, for example, a loop that does not have the local feature point. When processing a stroke that does not have the local feature point, the first detected point on the stroke is considered as if the starting point were the local feature point, and the above process (
3) and (4) are executed.

The local feature point table is sequentially supplied from the thin line figure 24 corresponding to the thin line figure shown in FIG. 2, while the contents of the local feature point table are read from the feature point memory 30. As a result, the microprocessor 31 stores the finally extracted global intersection points, global inflection points, global end points, and connection states in the coordinate point memory 3.
2 cents.

In addition to the above, in the present invention, in order to improve the conventional problem that occurs with the above-mentioned processing alone, the global inflection points extracted as described above are sequentially examined two by two, and the local inflection points are The degree of bias is detected, and a new global inflection point is generated according to the degree of bias. For this reason, the micro processor 31 is connected to a closed loop detection section 311 as shown in FIG.
, an eccentricity rate detection section 312, and a maximum distance difference calculation section 313. For details, please refer to page 2 in Figure 2.
The explanation will be given by taking the i-p2 as an example.

Here, a collection of local inflection points between global inflection points is regarded as a closed loop composed of various vectors with directions indicated by arrows, as shown in FIG. Generally, various types of closed loops such as those shown in FIGS. 6(a) to (e) can be considered, but here, FIG. 6(a),
Consider only the simple closed loop in (b). Figure 6 (C)
A closed loop like the one shown in FIG. 6 (al, (b)) is considered to be a somewhat accurate approximation, rather than a one-way bias as shown in FIG.

Identification of the type of closed loop can be performed by the sum of angular changes of vectors. The angle change is shown in Figure 7(a).
, +9 when turning right at a right angle as shown in (b)
0°, and -90'' when turning at a right angle to the left, then for simple closed loops such as those shown in Figure 6 (a) and (b), the sum of the angle changes is +360'' and -360°, respectively. Since the invention uses a four-way connection thinning process, as shown in FIG. 8, except for the angle change contributed by the global inflection point (a, b in the figure), (
As in b), only two angle changes occur: +90° and -90°. Further, the angle change contributed by the global inflection point is always ±180° as shown in FIG.

Therefore, in reality, to calculate the sum of angular changes, it is sufficient to check only the direction of angular change (that is, the sign of angular change), and detailed angle measurements are not required.

If it is determined that the closed loop is a simple closed loop from the sum of the angle changes, then the area inside the closed loop is determined. This can be easily carried out by the method shown in FIG. 9<8), (b). In other words, consider a coordinate system X'-Y' whose coordinate origin is the minimum X address and Y address of the start point and end point of the vectors that make up the closed loop. Find the direction of the vector with the minimum value. For example, when examining the direction of the vector having the maximum value of the Y' address, in the case of FIG. 9(a), the rightward vector b is detected, and in the case of FIG. 9(bl), the leftward vector h is detected. A vector having the same direction with respect to the X' axis as the direction of this detected vector (FIG. 9 (al)).

Regarding d), add the locus of the vector projected onto the X' wheel axis as the area, and move in the opposite direction with respect to the X' axis. Similarly, regarding the vector (f in Figure 9(a)),
The locus of the vector projected onto the X' wheel axis is drawn as the area. In other words, it becomes an addition form by adding a sign to the area.

In this way, the area S within the closed loop is obtained.

On the other hand, the perimeter of the closed loop can be simply determined by the sum of the vector sizes and lengths.

Through the above processing, we were able to find the area S within the closed loop and the perimeter of the closed loop for the simple closed loop.
From these two values, the following billion yen rate T is defined. That is, T=4π(S/L") For example, if the radius is r, the area S=πr2, the perimeter L
”=(2πr)t, so the billion yen rate T=1.On the other hand, in the case of a square, S=rt, LZ=(
4r)", so it is set as T-π/4. In other words, the billion yen rate T is related to the approximation accuracy when a collection of local inflection points is approximated by global inflection points, and the smaller T is, This is a good approximation.

Therefore, in this embodiment, the billion yen rate T is compared with a preset threshold value, and if the billion yen rate T is large, a new global inflection point is generated. The new global inflection point is selected as the local inflection point that is the furthest distance orthogonal from the vector between the global inflection points.

For example, if it is determined that the billion yen rate T for the closed loop shown in FIG. 5 is larger than the threshold value, the local inflection point A shown in FIG. 10 is newly generated as a global inflection point.

Once a new global inflection point is generated, the above process is further repeated until the billion yen rate T becomes smaller than the threshold.

Finally, the figure shown in FIG. 2 becomes as shown in FIG. 11.

The thick line in the figure is the approximated line.

Processing such as determination of a simple closed loop, extraction of area S, perimeter length, etc. is performed in the processor 31 shown in FIGS. 1(d) and 4.

〔Effect of the invention〕

According to the present invention, from freehand drawings drawn on paper etc.,
In particular, it becomes possible to grasp the intersection points in a thin line figure as a main feature, vectorize the screen information of the figure, and compress the information with a high degree of approximation without destroying the global features of the figure.

[Brief explanation of the drawing]

Figures 1 (a) to (d) are block diagrams of a graphic processing system as an embodiment of the present invention, Figure 2 is a diagram showing graphic data obtained by a conventional graphic processing system, and Figure 3 is a diagram of the present invention. FIG. 4 is a flowchart showing an overview of the processing of the processing system of the invention, FIG. 4 is a diagram showing a part of the processing of the microprocessor of FIG. 1(d), FIG. Figures (al~(C1 is an example of a closed loop according to the present invention, the seventh
The figure is an explanatory diagram of the angle change. Figure 8 is an explanatory diagram of the angle change contributed by the global inflection point. Figures 9 (a) and (bl are explanatory diagrams of area extraction. Figure 10 is the new global inflection point. FIG. 11 is an explanatory diagram of the inflection point generation process, and is a diagram showing graphic processing data obtained by the system of FIGS. , 2... Local feature point fine extraction section, 3... Information compression processing section. 1st 11a (C) Example diagram of local point feature point coordinate extraction section Fig. 1 (d) Information compression processing section Embodiment configuration diagram Figure 2 Diagram showing graphic data obtained by conventional graphic processing

Claims (1)

[Claims] A graphic processing system that compresses and expresses graphic information based on local feature points that define a drawn graphic as an intersection point, an end point, or a point on each closed loop. processing means; local feature point extraction means for extracting local feature points from the thin line figure obtained by the thinning processing section; traces at least one stroke of the stroke and extracts a local inflection point on the stroke until it reaches another local feature point, and also extracts a local inflection point between the local inflection points or between the local feature point and the local inflection point. a stroke tracing processing means for extracting the distance of a line element between the points; and generates a new global inflection point according to the degree of difference between the local feature point and a thin line figure based on the global inflection point. A graphic processing system characterized in that a stroke between the two points is expressed by at least the feature point coordinates and the global inflection point coordinates.