KR900002810B1 - Method for comperssing character or pictorial image data - Google Patents

Abstract

The method forms blocks and stores information for specifying the outline of a character as an approximation with sets of functional curves or straight lines to effect data compression. A character outline is developed on X- and Y-coordinates and is divided by split points into a set of blocks each defined by a single-valued function involving X as a variable. The character outline is approximated by a curve passing through sampling points established on each of the blocks. For each sampling point, the density of the sampling points in its vicinity is formed.

Description

Block setting method for character data compression

1 is a schematic explanatory diagram of a state in which a character outline is grasped as a set of plural blocks.

2 is a schematic illustration of the process of setting a sample point on any one block B according to the method of the present invention.

3 is a block diagram showing an embodiment of an apparatus suitable for implementing the method for setting a sample point of the present invention.

4 is a flowchart showing the operation of the configuration of FIG.

FIG. 5 is a diagram schematically showing an example in the case where a sample point setting method according to the configuration of FIG. 3 to FIG. 4 is performed.

6 is a diagram schematically showing an embodiment of a block setting method improved by the present invention.

FIG. 7 is a reference diagram showing an example in the case where an actual character outline is processed by using the sample point setting method and the block setting method in accordance with the present invention. FIG.

* Explanation of symbols for main parts of the drawings

B: block C: approximation curve

D, D ', Q: Split Point P: Sample Point

31: contour information storage unit 32: sample point coordinate storage unit

33: distance calculation unit 34: approximate curve calculation unit

35: area calculation part 36: evaluation amount calculation part

37: maximum evaluation amount detection unit 38: comparison unit

39: tolerance setting part 40: new sample shop calculation part

41: deviation calculation unit 42: maximum deviation position detection unit

"○": initial division point "raw": second division point

"△": Sample store

The present invention relates to a method for setting sample points and blocks when data compression is performed by approximating outlines of letters, images, etc. (hereinafter referred to as characters) to a set of function curves or straight lines.

First, setting of a sample point is demonstrated.

In the character data compression method in which the outline of the text is divided into sample points, and each divided outline segment is approximated by a function curve or a straight line, as few sample points as fast as possible are maintained within the range that does not impair the shape characteristics of the text outline. It is important to set.

Conventionally, as the simplest method for setting such a sample point, a method for obtaining a sample point by increasing the variable by unit amount has been used for the character outline developed on the X and Y coordinates.

However, this method is an optimal sample because the likelihood between the approximated curve (or straight line) and the actual character outline is not certain because the setting of the sample point depends only on the variable axis. It was difficult to set the point.

After the likelihood between the approximating curve (or straight line) and the actual character outline is obtained for the candidate surface point, the candidate sample point is moved by trial and error according to the likelihood, and at a point when a predetermined likelihood is obtained. There is also a method of using the candidate sample point as the set sample point, but this method requires a trial and error process in order to set one sample point.

The sample point setting method according to the present invention is to solve this problem, and to provide a reasonable method to set one sample point at a time by one likelihood comparison.

First, the preprocessing of the character outline which becomes the premise of sample point setting is demonstrated.

FIG. 1 is a schematic explanatory diagram of a state in which a character outline is divided into a plurality of blocks (B ₁ , B ₂ , B ₃ ...) And the character outline is grasped as a set of each block B. FIG. The raw table is the end point of each block.

In other words, as a preprocessing of sample point setting, a suitable dividing point is set on the character outline developed at the X and Y coordinates, and the block is obtained in advance by dividing the outline in the section of the monovalent function using x as a variable.

This block division can be generally handled by setting an arbitrary section in which the X coordinate of a series of outlines monotonously increases or decreases as a block.

In a more appropriate block setting method, as shown in FIG. 1, after obtaining the position where the x coordinate values on the series of contours become the extreme values, the position corresponding to the extreme values is determined by the initial dividing point D ( Table) and the second dividing point D 'according to a predetermined relationship on each contour block divided by this dividing point D, as described in detail with reference to the block setting method described later. By setting the " table), the section divided by each division point (D or D ') is set as the block B, respectively.

The physical meaning of each block B set in this way means that when a character image is reproduced on a lattice scan type character display device such as a cathode ray tube (CRT), each of the blocks B turns on a bright point scanned in the Y direction, Or a set of points to be turned off.

The method for setting a sample point according to the present invention relates to a method for setting a sample point for dividing the outline corresponding to the block into a plurality of contour segments for any one of the blocks B, which is set by the above preprocessing.

2 is a schematic explanatory diagram of a process of setting a sample point on any one block B according to the method according to the present invention, and FIG. 3 is a diagram illustrating a sample point setting method of the present invention. A schematic diagram illustrating one embodiment of a suitable apparatus.

In FIG. 3, reference numeral 31 denotes an outline information storage unit that stores coordinate information of outlines corresponding to any one block B shown in solid lines in FIG. 2 and outputs them as necessary. ) Is a sample point coordinate storage unit which sequentially stores the coordinate information of the sample point set by the method of the present invention and outputs it if necessary, and (33) is already set on the contour, that is, the sample point coordinate storage unit ( 32 is a distance calculation unit for calculating the distance L between adjacent sample points with respect to the preset sample points stored in 32), (34) is an approximate curve calculation unit to be described later, and (35) is cut out by adjacent sample points. An area calculating unit for calculating an area S surrounded by a true contour segment and an approximation curve connecting each sample point already set on the contour, and (36) is a distance L and an area calculated by the distance calculating unit 33; Calculated by the calculation unit 35 According to the enemy and S pyeonggaryang calculating section for calculating a pyeonggaryang ξ of S / L for each segment, 37, it is the maximum pyeonggaryang detector. (38) is a comparison unit for comparing the magnitude of the maximum value ξmax of the evaluation amount calculated by the evaluation amount calculation unit 36 and the predetermined allowable value ξ 'supplied from the allowable value setting unit 39, and (40) the comparison unit 38 In the case where the comparison result in the above is ξmax> ξ ', a new sample point calculation unit that calculates a new sample point according to a predetermined relationship, for example, comprises a deviation calculation unit 41 and a maximum deviation position detection unit 42.

The operation of the configuration of FIG. 3 will be described with reference to FIG. First, the coordinate information of both end points P ₁ and P ₂ (the _two end points correspond to the dividing point D or D 'of FIG. 1 obtained by the above-described preprocessing) of the block b of FIG. As the coordinate information, the sample point coordinate storage unit 32 is stored, and the initial condition of the operation is set.

When the setting of the initial condition is completed, the distance calculating unit calculates the distance L between the two sample points according to the coordinate information of the preset sample points P ₁ and P ₂ read out from the sample point coordinate memory 32. It calculates by (33) and outputs it to the evaluation quantity calculation part 36. FIG. In parallel with this, the approximation curve calculation unit 34 similarly calculates an approximation curve connecting the two sample points according to the coordinate information of the preset sample points P ₁ and P ₂ read out from the sample point coordinate storage unit 32. This is output to the area calculating part 35.

However, when the operation is started under the initial condition, since the preset sample points are two, the approximation curve becomes a straight line C ₁ as shown in FIG.

Next, the area S '+ S " = S shown in Fig. 2A is calculated based on the straight line C ₁ obtained by the approximation curve calculating unit 34 and the outline coordinate information stored in the outline information storage unit 31. It calculates in the part 35 and outputs it to the evaluation amount calculation part 36. FIG.

The evaluation amount calculator 36 calculates the evaluation amount ξ in S / L according to the distance L calculated by the distance calculator 33 and the area S calculated by the area calculator 35.

In this case, since there is one evaluation amount calculated, this value is supplied to the comparison unit 38 as the maximum value ξmax of the evaluation amount via the maximum evaluation amount detection unit 37.

Since the allowable value ξ 'is determined by the allowable value setting section 39 as the other input, the comparison section 38 operates the newly created sample point drawing section 40 when ξmax <ξ'.

The deviation calculation unit 41 constituting the newly-established head office calculation unit 40 is a contour coordinate stored in the curve (in this case, the straight line C ₁ ) and the contour information storage unit 31 obtained by the approximation curve calculation unit 34. According to the information, the deviation? Shown in FIG. 2 a is calculated over the entire period of the point P ₁ point P ₂ , and the positional information of the point P ₃ on the block B at which the value is maximum is the maximum deviation position point detecting unit 42. Obtain from Advantages and P ₃ is stored in the set sample point coordinate storage section 32 as the new sample point.

In this way, when the newly established sample point P ₃ is set on the outline of the block B, the configuration of FIG. 3 again executes the above operation. This operation is described below with reference to FIG. 2B. At this time, the new sample point P ₃ is updated with a predetermined sample.

That is, in this case, first, the distance calculation unit 38 calculates the distances L ₁ and L _{2 shown} in FIG. ₂ . In addition, the approximation curve C ₂ (indicated by the dotted line during the second period) passing through the predetermined sample points P ₁ , P ₂ , and P ₃ on the block B in parallel is obtained by the approximation curve calculation unit 34.

The area calculating part 35 calculates each contour segment cut out by the sample point, and the areas S ₁ and S ₂ (shown in diagonal lines) surrounded by the approximation curve C ₂ , respectively.

Next, the evaluation amounts ξ ₁ and ξ ₂ of S / L = ξ are determined by the evaluation amount calculation unit 36 according to the distances L ₁ , L ₂ and the areas S ₁ , S ₂ corresponding to the respective values.

In the case of FIG. 2B, since ξ ₁ <ξ ₂ , the value of ξ ₂ is supplied from the maximum evaluation amount detection unit 37 to the comparing unit 38 as the maximum value ξ max of the evaluation amount.

max로 되는 점 P₄가 신설 표본점으로 표본점좌표기억부(32)에 설정된다.Then, when ξmax (= ξ ₂ )> ξ ', the new sample point calculation unit 40 is operated again by the instruction from the comparing unit 38, and as described above with respect to the case of FIG.

The point P ₄ that is a max is set to the sample point coordinate storing section 32 in the new sample point.

Thus new sample point P ₄ is obtained after treatment conditions Since the second same as those shown in Figure 3 c also operation of the configuration itself will not be described below.

ξ'를 검지했을때, 임의의 1개의 블록 B에 관한 제 3 도 구성의 동작은 종료한다.In addition, in the process of obtaining a new sample point P ₁ , the comparison unit 38 measures ξmax.

When ξ 'is detected, the operation of the third FIG. configuration with respect to any one block B ends.

At this time, of course, is that the sample points P ₁ ~P ₁ stored in the sample point coordinate storage section 32 is set as a desired sample point on the outline of the block B.

X)²+(

y)²의 연산에 의하여 극히 간단히 구할 수 있다. 또 상기 면적 산출부(35)는 2개의 곡선의 y방향 차분을 x방향으로 적분하므로써 비교적 간단히 소망의 면적 S를 산출할 수 있다. 따라서 상기와 같이 S/L로서 부여되는 평가량 ξ의 값의 산출도 비교적 간단해진다. 그리고, 이 평가량 ξ는 면적 S를 길이 L로 나눔으로써 길이의 차원을 가지는 양으로 되기 때문에 면적 S를 형성하는 2개의 곡선(윤곽과 근사곡선)이 특이한 요철을 갖지 않는한 이들 양 곡선간의 이합(離合)관계를 표시하는 지표가 된다.Moreover, the distance L value in the said distance calculating part 33 is L = (

X) ² + (

y) Can be obtained very simply by operation of ² . The area calculating section 35 can calculate the desired area S relatively simply by integrating the y-direction differences of the two curves in the x direction. Therefore, the calculation of the value of the evaluation amount ξ given as S / L as described above is also relatively simple. Since this evaluation amount ξ is obtained by dividing the area S by the length L to obtain a dimension having a length dimension, the difference between the two curves (contour and approximation curve) forming the area S does not have an unusual unevenness ( It is an indicator of the relationship.

4 is a flowchart illustrating an operation process of the configuration of FIG. 3.

Next, a block setting method for making the sample point setting method according to the configuration of FIG. 3 to FIG. 4 more highly effective in relation to the case where the character outline has the above-mentioned unevenness will be described.

That is, when the sample point setting method according to FIGS. 3 to 4 is implemented, the sample point P is close to the portion where the curvature radius of the character outline is small as shown in FIG. 5. In this case, the improvement of the data compression ratio depending on the sample score occurs.

Therefore, the improved block setting method described below takes advantage of the dense property of the sample point and uses it, and the second splitting point (the dividing point in FIG. 1) at the portion where the set density of the sample point exceeds a predetermined threshold value. D ') is set newly, and the block is divided into two, and for each of the divided blocks, the sample point setting in the entire character outline is set again by setting the sample point according to the configuration of FIG. 3 again. It was to decrease.

An embodiment of this improved block setting method will now be described with reference to FIG.

In Fig. 6, P ₁ and P _n are end points of any one block, and P _{1-1 to} P _{1 + 3} are sample points set by applying the sample point setting method to block B.

의 값을 i=2∼(n~3)의 모든 경우에 대하여 산출하고, 각각 임계치 ρ'와 비교한다.In implementing the improved block setting method, for example, first,

The value of is computed for all cases of i = 2-(n-3), and is compared with threshold value rho ', respectively.

Then, the case where ρ ₁ <ρ 'is extracted as the sample point density interval.

에 대하여 가장 떨어진점 Q를 구한다.Next, a straight line in this sample point dense section, that is, the section [P _1-1,: P _{1 + 3} ] on block B

Find the lowest point Q with respect to.

와 직선

로 구성되는 각 (P_1-1, Q, P₁₊₃)=θ_q를 구한 후 이것을 임계치 θ'와 비교하고, θ_q〈θ'가 되는 경우의 점 Q를 새로운 분할점 D'로 하고, 1개의 블록을 2분 한다.And straight

And straight

_Calculate the angle (P _1-1 , Q, P _{1 + 3} ) = θ _q and compare it with the threshold θ ', and let point Q be the new splitting point D' when θ _q <θ '. 1 block for 2 minutes.

The improved block setting method is as described above. By applying the method of setting the sample point of FIG. 3 again to each of the two divided blocks, the sample score is decreased, especially near the singular point such as the straight intersection point of the character outline or the so-called tip of the `` spiking ''. This contributes to the data compression effect.

An example of the case where the actual character outline is processed according to the sample point setting method and the block setting method described in detail above is shown with reference to FIG.

In FIG. 7, the "○" table is an initial division point described with reference to FIG. 1, the "raw" table is a second division point obtained by an open block setting method, and the "△" table is the second division point. Sample point obtained after setting the split point of. The number of these points shown in the drawings for reference is 40, 9 and 99.

In the present invention, as described in detail above, a plurality of samples obtained by grasping a character outline that is typically developed on X and Y coordinates as a set of blocks formed by a monovalent function using x as a variable, and again set on each block. In the problem data compression method in which the character outline is approximated by a curve passing through a point, it is used to take advantage of the feature that the sample point is close to the portion where the radius of curvature of the character outline is small. Calculating the density of the multiple sample points set on the block for each sample point, and then setting a new dividing point in the dense part when the sample point closeness where the density exceeds a predetermined threshold is detected; The division point is used to divide and set the arbitrary one block into two blocks. In this way, the present invention reduces the number of sample point settings in the entire character outline, and provides the block setting method as the character data compression method is more effective.

Claims (2)

Grasping the outline of the letter outlined on the X, Y coordinates as a set of blocks formed by a monovalent function using x as a variable, and again by the curve passing through the multiple sample points set on each block A character data compression method for approximating an outline, comprising the steps of: calculating the density of sample points set on any one block for each sample point; When a sample point cluster having a density exceeding a predetermined threshold is detected, a new splitting point is set in the cluster, and the arbitrary one block is divided into two blocks by the splitting point. A method for solving problem data compression, characterized by setting a block for character data compression. The method of claim 1, wherein when P ₁ (where i = 1 to n and P ₁ and P _n are the starting point and the ending point of the block), P _i and P _{i are set. +1} + P _{i + 1} , P _{i + 2} = ρ _i {where i = 2 to (n to 3)} is less than the predetermined threshold ρ '

Find the furthest point Q for, and again, straight

And straight

Character data characterized in that when the angle (P _i-1 , Q, P _{i + 3} ) = θ _q constituted by is smaller than a predetermined value θ ', the point Q is set as a desired dividing point. Compression block setting method.

Images