JPH07152920A - Graphic processor and its method - Google Patents

Abstract

PURPOSE:To solve an acute end and an overlap problem and to enable high- speed software processing. CONSTITUTION:An instruction means 1 instructs the kind, size, and modification of a graphic to be processed. Then outline data on the graphic instructed by the instruction means 1 are supplied from an outline data memory 2 to an outline raster scanning means 3. The outline data are so set that the direction of external loop data is opposite to the direction of internal loop data. Then the outline raster scanning means 3 makes a raster scan. For this raster scan, one outline point forming one outline is determined on one line and generated in a painting-out buffer 5. Further, a painting-out means 4 performs an additive subtractive painting-out process and the value of respective pixels of the painting-out buffer 5 are increased or decreased; when the value becomes larger than 1, the pixels are painted out on each line. Lastly, the graphic stored in the painting-out buffer 5 is displayed at a display means 6.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a graphic processing apparatus and method, and more particularly, when converting outline vector data indicating the outline of a graphic such as a character into dot data, the inside and outside of the outline and the graphic are painted. The present invention relates to a crushing graphic processing apparatus and method.

[0002]

2. Description of the Related Art Conventionally, in order to display a high-quality character even when it is enlarged or reduced, the outline of the character is held as outline data, and the character drawn by converting the outline data into dot data. The method of filling the inside or outside of the contour of the is adopted.

As a method of enlarging / reducing coded characters and displaying them on an output device such as a cathode ray tube,
Japanese Patent Publication No. 53-15624 “Character Generation Method” and Japanese Patent Publication No. 53-410
No. 17, “Method of generating high-quality characters, etc.” is known.
However, these methods have a problem that the character is not filled correctly when the sharp edge portion of the character outline is 1 dot, and in the following patent applications, etc., the method for solving this sharp edge problem is solved. Have been proposed.

As a method for solving the sharp edge problem and filling out outline vector data of characters and figures, there is disclosed in
The "image processing device" of 2-59872, the "graphic filling method" of JP-A-4-684, and the "outline data drawing device" of JP-A-3-233689 are proposed. And Japanese Patent Laid-Open No.
The 2-59872 "Image processing device" is a relatively excellent method that solves the sharp-edge problem by a simple method suitable for software processing. By performing the above, the filling operation is performed with a simple structure while avoiding the problem of sharp edges. However, there is a problem in that the processing speed is lowered because it is necessary to locally perform inversion processing a large number of times for one scan line. Therefore, in this proposal, the filling processing area is divided into two in the vicinity of the center line of the figure,
By changing the filling direction in each area, the degree to which the inversion processing locally overlaps is reduced, but even in this case, the problem that the processing speed decreases remains.

Further, as a problem of filling, there is a problem of overlapping of contours. This duplication problem will be described below by taking a character thickness modification, which is one of the modifications of outline characters, as an example.
For example, when the katakana "mu" is defined by an outline vector as shown in FIG. 10A, an example of a method of performing thickness transformation is, as shown in FIG. To move the vertex coordinates in a direction that bisects the angle formed by the side. At this time, the contour lines may collide or overlap as shown by a thick arrow in FIG. 7B. In this case, the method of Japanese Patent Laid-Open No. 2-59872 has the following problems. Occur.

That is, FIGS. 11 (A) to 11 (E) show enlarged views of a portion where the contour lines overlap each other, and the reverse filling procedure for filling the inside of the character surrounded by the contour lines is shown. To explain, first, as shown in FIG. 7A, the right side of the leftmost contour line is inverted to black, and then the right side of the adjacent contour line is made white as shown in FIG. Invert to blacken only the area enclosed by the outline. However, when the same processing is performed, as shown in (C) and (D) of FIG.
Since the overlapped portion of the portion surrounded by the contour line is inverted again, a white hollow portion occurs even though it is surrounded by the contour line as shown in FIG. There was something that happened.

Japanese Patent Laid-Open No. 4-684 “Graphic Filling Method”
Solves the duplication problem caused by scaling by expressing a pixel (pixel) by a combination bit of a contour flag and a direction bit, and rewriting while referring to this pixel at the time of contour drawing and at the time of filling. This is what we are trying to do (note that this method does contouring and filling separately). In particular, it is possible to deal with the disappearance of the stroke due to the degeneracy of the parallel contours accompanying the character reduction as shown in FIG.

However, when the contour drawing is performed by this method, there is always a case where the rule for updating the direction information of the pixel is omitted after the start of the contour drawing. For example, in this update rule, since the direction value for the current pixel is determined by the data of the previous and next pixels, the processing immediately after the start point of the contour is not determined (FIG. 13 (A)). Further, if a point where the Y coordinate value does not change continues immediately after the start point of the contour, indefinite points may continuously occur ((B) in the same figure). Therefore, in addition to the procedure disclosed in Japanese Patent Laid-Open No. 4-684, it is necessary to search for an indefinite point and determine it again after one round of contour drawing.

Further, although the process from the creation of the most important direction information to the filling is not clearly described in this proposal, in this case, the filling as shown in FIG. 14 (A) is not performed inside. When vector data such as that shown in FIG.
FIG. 16 shows the outline drawing when the direction is determined and the indefinite point is determined again by the update rule. The black circles are points where the contour bits are set, and the points where no numerical value is written are points where the values are in the initial state (= 0).

When the values for each pixel are added in the horizontal direction, the result shown in FIG. 17 is obtained. On the other hand, by performing a condition judgment of zero non-zero even odd number and performing a logical sum with the contour information, the figure shown in FIG.
The results as shown in 5 (B) to (E) are obtained. At this time, as shown in FIG. 14 (A), in order to accurately fill a figure having a portion which is not filled inside, an odd condition that can obtain the result shown in FIG. 15 (E) is set. Will be adopted.

Here, a problem when the contour lines overlap as described in the thickness modification example will be described.
If there is vector data as shown in FIG. 18B for a figure as shown in FIG. 18A (a diagram schematically showing a situation caused by a change in thickness), the previous example Similarly, when the direction is determined by the update rule and the indefinite point is re-determined and the contour is drawn, the result is as shown in FIG.

Then, when values are added pixel by pixel in the horizontal direction, the result shown in FIG. 20 is obtained. Here, if the condition to be adopted is the odd number as in the above case, the value of the overlapping portion becomes -2, which is not an odd number, so that the omission occurs at the overlapping portion. On the other hand, if non-zero is used as the condition, the inside of the inner loop, which originally does not need to be filled, is filled, which causes an essential problem. Furthermore, including the problem of indefinite point processing that I pointed out, this proposal has a lot of judgment processing based on special rules,
Also, since there are many branches to special processing, it takes a very long processing time when software processing is performed by using the apparatus having the configuration shown in the embodiment of this proposal.

The "outline data drawing device" of JP-A-3-233689 is a method for individually performing contour drawing and filling based on a special rule similar to that of JP-A-2-59872 described above. A detailed explanation is given on the countermeasure for the blanking problem of the overlapping portion of the contour line, which is a problem, but the rule used is more complicated than the one that introduces the state transition. Therefore, although one of the objectives is to speed up hardware, MPU (microprocessing un
It) took a much longer time to process software.

[0014]

In the method of performing contour drawing and filling separately, a sharp edge is dealt with by not storing the contour as a simple set of dots but by storing the property of each part of the contour in some form. Had to do. On the other hand, with the method disclosed in Japanese Patent Laid-Open No. 2-59872, the sharp tip problem can be avoided, but the duplication problem cannot be avoided.

Further, in the conventional method, if it is attempted to avoid both the sharp edge problem and the duplication problem, it takes a long time to process the software, which is not very practical. SUMMARY OF THE INVENTION It is therefore an object of the present invention to solve the above-mentioned problems caused by sharp edges and contour overlap by a simple method, and to provide a graphic processing apparatus and method capable of high-speed processing when performing software processing by an MPU. And

[0016]

As means for achieving the above-mentioned object, an instruction means for instructing at least the type and size of a figure, a direction of outer loop data indicating an outer contour in which a loop is a figure, and a loop Outline data storage means for storing the outline data set so that the direction of the inner loop data showing the inner contour of which the outside is a figure is opposite, and the outline data of the figure indicated by the instructing means is the above-mentioned. Contour rasterizing means which is supplied from the outline data storing means and rasterizes the contour line indicated by the outline data by determining one contour point for each line and the contour line determined by the contour rasterizing means. Is divided by the contour line in advance depending on whether the line processing direction is an increasing direction or a decreasing direction with respect to the vertical direction. A graphics processing device comprising: a filling means for incrementing or decrementing the value of each pixel on the one direction side set for each line; and a filling buffer for storing a graphic filled by the filling means, And, a method of processing a figure that fills the inside or outside of this figure when converting outline data showing the contour of the figure into bitmap data, and the direction of the outer loop data showing the outer contour where the inside of the loop becomes the figure. The contour line indicated by the outline data, which is set so that the direction of the inner loop data indicating the inner contour outside the loop is a direction, is set to 1 for each line.
The value of each pixel on one side in a predetermined direction divided by the contour line depending on whether the contour line is determined and rasterized and the contour line increases or decreases with respect to the line processing direction and the vertical direction. Is to be incremented or decremented line by line.

[0017]

By performing contour drawing and filling in parallel, the sharp edge problem can be dealt with without memorizing the nature of the contour. In addition, whether the outer loop data and the inner loop data stored in the outline data are clearly distinguished by the direction of the loop, and one line is added / subtracted for each contour point to see if the value of each pixel is filled By determining whether or not it is possible to prevent the omission of the overlapping portion. Furthermore, by processing multiple pixels in a word length that is easy for the MPU to process, by using a simple algorithm without special judgment,
The processing speed can be improved.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the graphic processing apparatus and method of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of the graphic processing apparatus of the present invention. In the figure, the instruction means 1 is for instructing the type of graphic to be processed (which graphic is to be processed), the size such as enlargement and reduction, and the modification of the thickness and shape. Then, the outline data of the figure instructed by the instructing means 1 is supplied from the outline data memory (outline data storing means) 2 to the contour rasterizing means 3. The outline data of each figure stored in the outline data memory 2 includes the direction of the outer loop data indicating the outer contour for judging the inside of the loop as the figure and the direction of the inner loop data indicating the inner contour for judging the outside of the loop as the figure. The data of the vertices and the data of the type of the line are arranged in the opposite direction. Then, the contour rasterizing means 3 rasterizes the outline data to create a contour. In this rasterization, as will be described later, the contour points forming the contour line connecting the vertices are set to 1 (in the filling buffer 5).
One for each line is determined, and drawing is performed in the filling buffer 5 via the filling means 4.
Further, the filling means 4 performs addition and subtraction filling described later, increments or decrements the value of each pixel (pixel) of the figure drawn in the filling buffer 5, and fills the pixels that are below or above a certain value. The figure is filled with the selected pixels. Finally, by displaying the figure stored in the filling buffer 5 on the display means 6, the figure instructed by the instructing means 1 can be displayed.

FIG. 2 shows a concrete example of the graphic processing apparatus of the present invention. Explaining in comparison with the configuration diagram shown in FIG. 1, the keyboard 1a corresponds to the instruction means 1 for instructing the type, size, and decoration of a graphic to be processed.
MPU (microprocessing unit) 7 is a program RO
According to the program stored in M8, the outline data supplied from the outline data ROM 2a is operated, and the filling buffer R such as VRAM is filled.
This is a circuit for outputting to the AM 5a to draw and fill a figure. The outline data ROM 2a corresponds to the outline data memory 2 of FIG. 1, and the MPU 7 and the program ROM 8 include the contour rasterizing means 3 and the filling means 4.
Is equivalent to. The fill buffer RAM 5a corresponds to the fill buffer 5. The figure stored in the fill buffer RAM 5a is converted into an analog video signal by the DA converter 9, and the display means 6 such as a CRT is displayed.
Is output to.

Next, the method of rasterizing and filling the outline of the figure, which is the main part of the present invention, will be specifically described. In the drawings from FIG. 3 onward, the two-dimensional image plane generated in the fill buffer 5 is represented by XY coordinates, the increasing direction of the X coordinate is the direction in which the filling process proceeds, and the increasing direction of the Y coordinate is a character or figure. The upward direction and the increasing direction of the X coordinate are expressed as the right direction of the character or figure. Further, the outline data stored in the outline data memory 2 and showing the outline of the character to be filled is
Although general outline data expressed by a straight line or an approximate curve may be used, the arrangement direction of the data needs to be expressed in the outer loop and the inner loop in opposite directions to clearly distinguish them. Since the outline data is generally associated with the expansion processing system, there is no problem even with such data.

(1) First, the creation of the contour will be described. To create a contour from outline data,
An algorithm generally called a DDA (Digital Differential Analyzer) algorithm is used. The DDA algorithm is described in various domestic publicly known documents and, for example, is introduced in detail in "Practical computer graphics--basic procedures and applications--" supervised by Fujio Yamaguchi. In order to rasterize a straight line, the DDA algorithm expresses the solution of the differential equation indicating the straight line in the form of a recurrence formula, and determines the contour point while increasing the value of X or Y by 1 pixel. is there. And DD
Among the A algorithms, the integer Bresenham algorithm is convenient for use in a general digital processing system because it does not use a decimal value and does not perform division. It is also widely used because it can plot a linear image with few errors.

This Bresenham generalized algorithm is C
It is generalized in consideration that the brightness of a straight line drawn on RT or the like is visually constant regardless of the direction of the straight line. Then, the direction of the straight line is considered as an octant on the XY plane, and the second and third, fourth and fifth, sixth and seventh, eighth and first
With respect to each set of octants, the four types are associated with each other: whether X is selected as a variable of the algorithm, Y is selected, whether each selected variable is incremented, or decremented.

In this embodiment, the Bresenham algorithm is used, but the purpose thereof is not to make the brightness of the drawn straight line visually constant regardless of the direction of the straight line, but to start the filling process. The point is to decide.
Considering all four types of processing, the number of filling start points determined once is not fixed to one when viewed from the filling processing direction, which causes inconvenience. Therefore, in this embodiment, regardless of the direction of the straight line vector. Instead, only Y is used as a variable, and only the increment is performed (the processing content corresponding to the pair of the second and third octants of the integer type Bresenham algorithm), thereby defining the starting point.

Here, the relation between the DDA algorithm and the filling will be described. Figure 3 (A) is an integer type Bresenha
m This is an example of the processing result corresponding to the first octant of the generalized algorithm, and FIG. 6B is an example of the processing result for the second octant. Then, in the same figure (A), the contour point is determined while incrementing by 1 in the positive X direction, and in the same figure (B), the contour point is determined.
The contour points are determined by incrementing them one by one in the positive direction. When the filling process is performed in the positive X direction with the contour point thus determined as the starting point of the filling process, in FIG. 9A, one Y coordinate value (same line) forming a single contour line is set. On the other hand, since there are a plurality of contour points, there are a plurality of filling start points, which causes inconvenience. As a result, not only the processes are duplicated, but also errors may occur in the results of inversion and addition / subtraction.

On the other hand, in FIG. 7B, since there is only one contour point for one Y coordinate value forming a single contour line, the filling start point is also determined as one. Therefore, no inconvenience occurs. Therefore, in the present embodiment, a determination is made as to which quadrant the outline data fits in, and the DDA algorithm is used so that the variable is only the Y coordinate in correspondence with the set of the second and third octants. Use by transforming. It should be noted that the correspondence does not always make the brightness constant, but since the contour point obtained as a result of the processing is used as a starting point for filling, a change in the brightness of the contour or the like does not pose a problem.

(2) Next, the filling will be described focusing on the pointed tip problem. The filling method according to the present embodiment is the modification D.
A contour line is drawn by determining and rasterizing one contour point for each line using the DA algorithm, and when this contour line is in a direction increasing with respect to the line processing direction and the vertical direction, A pixel that increments (increases) the value of each pixel on the right side separated by the contour line for each line, and decrements (decreases) when it is in the decreasing direction, and fills the pixels whose value is 1 or more. It is what

Specifically, first, each pixel is represented by an integer value of several bits, and the initial value is set to 0. Next, the modified DD
In the A algorithm, each time the filling start point is determined, addition or subtraction of each pixel value is performed in the increasing direction of the X coordinate value. Then, when performing the modified DDA algorithm process, if the Y coordinate value of the starting point of the straight line is smaller than the Y coordinate value of the ending point, the filling starting point is determined from the starting point to the ending point. At that time, the operation for each pixel is an increment. If the Y coordinate value of the start point of the straight line is larger than the Y coordinate value of the end point, the filling start point is determined from the end point toward the start point. On the contrary, the operation for each pixel is decrement.

Next, description will be given of a case where a figure including a sharp edge is filled by this method. 3 points A (0,
Consider a triangle consisting of 0), B (4,8) and C (7,0). Since this triangle is an outer loop, it is defined clockwise. First, as shown in FIG. 4, the contour points corresponding to the line segment AB are plotted and filled. Contour points are set at each point indicated by a black circle in the direction of the arrow in the figure. Every time one contour point is determined, addition and subtraction processing is performed on the value of that contour point and all the points to the right of that contour point (white circles). . in this case,
Since the line segment is in the Y coordinate increasing direction, addition processing is performed, and the value of each point on the right side becomes +1.

Next, as shown in FIG. 5, contour point determination and filling processing corresponding to the line segment BC are performed. Although the contour line is defined in the direction from the point B to the point C, it is reversed and drawn in the direction of the arrow (direction from the point C to the point B). The contour point is determined by each point of the black triangle, and every time one contour point is determined, the value of the contour point and all the points (white triangle points) on the right side of the contour point are added and subtracted. In this case, since the line segment is reversed in the Y coordinate increasing direction, the subtraction process is performed. As a result, the value of each point becomes +1 only at the black circle points shown in FIG. 6, and becomes 0 at the other points. Further, since the Y coordinate values of the point C and the point A are equal, it is not necessary to process the line segment CA.

By performing the filling in this way, there is no inconvenience such as when the figure including the sharp edge is filled so that the sharp edge has a tail on the right side, the accurate filling can be performed. it can. Here, the loop condition of the DDA algorithm for one line segment is illustrated as Y variable <Y coordinate value of the end point, but the condition is Y variable ≦
The result is the same even if the Y coordinate value of the end point is used.

(3) Further, the filling problem will be described focusing on the duplication problem. A description will be given of a process of overlapping graphics that occurs when the katakana “mu” defined by the outline vector shown in FIG. 10A used in the conventional example is thickened as shown in FIG. In addition, the same figure (B)
In FIG. 7A to FIG. 7E, description will be given by enlarging a portion in which the outline of the outline indicated by a thick arrow has occurred. Here, as in the case of (2), a method combining the modified DDA algorithm and addition / subtraction filling is used.

First, when the line segment A shown in FIG. 7A is processed, since the line segment A is in the Y coordinate value increasing direction, the left side of this line segment A remains the initial value 0 and the line segment A +1 on the right
Becomes Next, as shown in FIG. 6B, when the line segment B is processed, the line segment B is in the direction of decreasing the Y coordinate value.
Between the line segment A and the line segment B remains +1 and the right side of the line segment B is deducted 0.

Further, as shown in FIG.
Is performed, the line segment C is in the Y coordinate value increasing direction, so all the pixels on the right side of the line segment C are set to +1. As a result, on the right side of the line segment C, the part between the line segment A and the line segment B (the part between the line segment C and the line segment B) becomes +2, and the other part (the line segment B and the line segment B It becomes +1 in the part on the right side of the line segment C). Also, line segment A on the left side of line segment C
The part sandwiched by the line segment B and the line segment B (the part on the left side of the line segment C and the line segment B) remains +1 and the other part (the part between the line segment B and the line segment C) remains. It remains 0.

Then, as shown in FIG.
As for the line segment C, all pixels on the right side of the line segment D are set to +1 because they are in the Y coordinate value increasing direction. The part sandwiched by the line segment B is set to +2, and the other parts are set to +1. Finally, as shown in FIG.
When the process is performed on the line segment E, since the line segment E is in the direction of decreasing the Y coordinate value, all the pixels on the right side of the line segment E are set to -1, and the value is deducted. Then, by setting the pixel having a positive value (1 or more) obtained by such processing as the pixel to be filled, the correct filling result can be obtained even in the portion where the contours overlap.

Therefore, by using the method combining the modified DDA algorithm and the addition / subtraction filling,
It can solve both the sharp point problem and the overlap problem.
Further, in the above-described embodiment, the outer loop data indicating the outer contour in which the inside of the loop is a figure has been described. By holding the outline data so that it is in the opposite direction to the data, correct filling can be performed by exactly the same processing.

(4) Finally, a method of speeding up the filling will be described. Here, a method for speeding up the software processing especially using the MPU will be described. In the filling method of the present invention described above, it is necessary to represent one pixel with a plurality of bits in order to perform addition / subtraction processing on each pixel. How many bits are actually required depends on the complexity of the character graphic, but when used for Japanese outline fonts, about 4 bits is usually sufficient. On the other hand, the processing unit of data generally handled by the MPU is a power of 2, and when a relatively large amount of data is processed, the MPU having a data bus width of 16 bits or 32 bits is usually used.

Here, as an example of efficiently performing the filling process by combining a plurality of pixels, consider the process of data represented by 4 bits / pixel using a 16-bit accessible buffer memory. In this case, the pixel arrangement and the memory arrangement can be associated with each other as shown in FIGS. Note that FIG. 9A is a diagram showing a pixel arrangement when a character image is expressed on the buffer memory, and FIG. 9B is an address arrangement diagram of the buffer memory as seen from the MPU. Since 4-bit pixels are arranged in a 16-bit memory for 1 word, 4 pixels are arranged in 1 word.

Here, the filling start point (contour point) determined by the DDA algorithm or the like is four parts of bit 15-12, bit 11-8, bit 7-4, and bit 3-0 in the memory layout. Is a data word associated with either of the above. On the other hand, in the data word corresponding to the pixel arrangement upper right side of the data word associated with the filling start point, all bit parts are incremented or decremented.

Therefore, as shown in FIG. 9, the filling of the starting point is divided into the above four cases, but all the data words corresponding to the right side of the pixel arrangement in the pixel arrangement correspond to the filling starting point in bits 15-12. It suffices to perform the same processing as the case of being attached. And the above 4
Assuming that the number of pixels for one line in the X-axis direction is an integer multiple of the number of pixels (4) represented by one data word (in the case as shown in FIG. 9) in the case of two cases, Corresponding can be made according to the remainder when the X coordinate value of the determined painting start point is divided by the number of pixels represented by one data word (in the figure, the remainder is 0, 1, 2,
3 is a, b, c, d ... An, bn, c, respectively.
(corresponding to each row of n and dn).

As described above, the four cases are associated with each other to increment or decrement the filling start point and the pixels on the right side of the filling start point, and the pixels which finally become 1 or more (or 0 or less) are filled. It should be treated as if

Next, an actual representation method of the value of a pixel and an actual addition / subtraction method will be described. As described above, the value of a pixel may be processed by a normal integer expression, but by devising the expression method, the above combination processing can be performed efficiently when processing by software. it can. It should be noted that this method enables high-speed processing even when it is configured by hardware. Normally, when a signed integer is expressed in binary, it is often expressed in 2's complement notation, but since it can also be expressed in the form of unsigned binary coded integer with an offset, the latter code is used here. None A case of expressing in the form of a binary-coded integer will be described as an example (note that the numbers to be added and subtracted are expressed in 2's complement).

When an unsigned binary coded integer with offset is represented by n bits, the value of 0 is all set except for the most significant bit, and the value of 2 ⁿ / 2 is all clear,-((2 ⁿ / 2 ) -1) is represented by a bit code in which all values are set. An example of a 4-bit code is as follows.

[0042]

To perform addition and subtraction (increment and decrement) on this value, it is established by performing an operation of increment: +0001 decrement: +1111 (however, carry is ignored). , And the added value is the same even when the augend is expressed in 2's complement).

A case where this arithmetic processing is performed by combining a plurality of pixels will be described by taking the cases of FIGS. 8A, 8B and 9 as an example. When incrementing for a combination of 4 pixels, the added value is divided into the following four cases depending on the position of the bit part corresponding to the filling start point as described above.

Fill start Add value (binary representation 16 bits) Bits 15-12: 0001 0001 0001 0001 Bits 11-8: 0000 0001 0001 0001 Bits 7-4: 0000 0000 0001 0001 Bits 3-0: 0000 0000 0000 0001

On the other hand, in the case of decrementing, the carry propagates to the upper bits, so it is necessary to pay attention to this processing. For example, when decrementing both of the two initialized pixels, the initial value: ± 0 0111 0111 Addition value: -1 1110 Carry 1111 Addition result: 1000 ← 1000. Except for the pixel portion, it is necessary that the added value is not all set (-1 in 2's complement expression) and the least significant bit is clear (-2 in 2's complement expression).

Further, when the position of the bit part corresponding to the filling start point is not at the highest position, the added value of the bit part corresponding to the left side of the start point (without decrement) is
It must be all-set (-1 in 2's complement representation) to offset the carry. For example, when decrementing only the lower pixel of the two initialized pixels, the initial value: ± 0, ± 0 0111 0111 additional value: ± 0, -1111 carry 1111 additional result: 0111 ← 1000 To do so.

Therefore, when decrementing for a combination of 4 pixels, the added value is divided into the following four cases depending on the position of the bit portion corresponding to the filling start point, and the added value corresponding to each is used. This enables correct decrement.

Fill start Add value (binary representation 16 bits) Bits 15-12: 1110 1110 1110 1111 bits 11-8: 1111 1110 1110 1111 bits 7-4: 1111 1111 1110 1111 bits 3-0: 1111 1111 1111 1111

Although the increment and decrement in the case of 4 pixels have been described above, the present invention can similarly obtain an addition value and perform addition and subtraction in the case of other numbers of pixels and 32 bits. Since it is possible to perform increment or decrement by simply adding the added value, high-speed arithmetic processing is possible, and moreover, it is sufficient to determine the pixel to be filled after obtaining the final value, It is not necessary to perform the inversion process, and the filling process can be performed at high speed.

[0051]

The graphic processing apparatus and method of the present invention are
Since the contour drawing and the filling are performed in parallel, it is possible to accurately fill the sharp end without memorizing the nature of the contour.

Further, after the direction of the outer loop data and the direction of the inner loop data stored in the outline data are reversed to be clearly distinguished, one line is added / subtracted for each contour point to obtain each line. Since it is possible to judge whether the pixel is filled or not based on whether the pixel value is positive or negative,
It is possible to prevent the omission of the overlapping portion. Further, it is possible to perform accurate filling by the same processing without distinguishing the outer loop data and the inner loop data.

Furthermore, since a simple algorithm is used in a state where a plurality of pixels are grouped into a word length that is easy for the MPU to process, and a simple algorithm is used, high-speed software processing is possible. In addition, the memory can be used with high efficiency. Further, even when the processing is performed by hardware, it can be realized with a simple configuration.

Further, by simultaneously handling a plurality of pixels, a function capable of processing a data bit number which is multiple times the hardware data bus width, which is partially implemented in recent MPUs, by a single instruction is effective. There is an effect that it can be used.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an embodiment of a graphic processing device of the present invention.

FIG. 2 is a configuration diagram showing a specific example of a graphic processing device of the present invention.

FIG. 3A is a diagram showing an example of a processing result corresponding to the first octant of the integer type Bresenham generalized algorithm; FIG.
FIG. 9B is also a diagram showing an example of a processing result corresponding to the second octant.

FIG. 4 is a diagram for explaining filling of a figure including a sharp end according to the present invention.

FIG. 5 is a diagram for explaining filling of a figure including a sharp end according to the present invention.

FIG. 6 is a diagram for explaining filling of a figure including a sharp end according to the present invention.

7 (A) to 7 (E) are diagrams for explaining the overlapping problem of filling in the present invention.

FIG. 8A is a diagram showing a pixel arrangement when a character image is expressed on a buffer memory in an embodiment of the present invention, and FIG. 8B is a diagram showing an address arrangement of the buffer memory as seen from the MPU. is there.

FIG. 9 is a diagram showing a pixel arrangement for explaining case classification of filling of a starting point in an embodiment of the present invention.

10A is a diagram showing an example of a character, and FIG. 10B is a diagram showing an example in which the thickness of the character is modified.

11A to 11E are diagrams for explaining a conventional overlapping problem of filling.

FIG. 12 is a diagram for explaining a conventional example.

13A and 13B are diagrams for explaining the problems of the conventional example.

14A is a diagram showing an example of a graphic, and FIG. 14B is a diagram showing an example of a vector thereof.

FIG. 15A is a diagram showing an example of a figure, and FIGS.
FIG. 8 is a diagram showing an example of a result of filling the figure.

16 is a diagram showing an example in which contour drawing of the graphic of FIG. 14 is performed.

FIG. 17 is a diagram showing an example in which contour drawing of the figure in FIG. 14 is performed.

18A is a diagram showing an example of a graphic, and FIG. 18B is a diagram showing an example of its vector.

FIG. 19 is a diagram showing an example in which contour drawing of the figure of FIG. 18 is performed.

20 is a diagram showing an example in which contour drawing of the figure in FIG. 18 is performed.

[Explanation of symbols]

1 Indication Means 1a Keyboard 2 Outline Data Memory (Outline Data Storage Means) 2a Outline Data ROM 3 Outline Rasterization Means 4 Filling Means 5 Filling Buffer 5a Filling Buffer RAM 6 Display Means 7 MPU (microprocessing unit) 8 Program ROM 9 DA Converter

Claims (4)

[Claims] 1. An instruction means for instructing at least the type and size of a figure, a direction of outer loop data showing an outer contour in which the inside of the loop is a figure, and a direction of inner loop data showing an inner contour in which the outside of the loop is a figure. And outline data storage means for storing outline data set so as to be in the opposite direction, and the outline data of the graphic indicated by the instruction means is supplied from the outline data storage means, and the outline indicated by the outline data Contour rasterizing means for deciding and rasterizing a line by determining one contour point for each line, and whether the contour line determined by this contour rasterizing means increases in the line processing direction and the vertical direction. For each line, the value of each pixel on the predetermined one direction side that is divided by the contour line depending on the decreasing direction is A fill means for incrementing or decrementing, graphics processing apparatus being characterized in that a fill buffer storing graphic to be filled by the fill means. 2. A figure processing method for filling the inside or outside of a figure when converting outline data showing the figure outline into bitmap data, wherein outer loop data showing an outer contour in which a loop is a figure. And the contour line indicated by the outline data set so that the direction of the inner loop data showing the inner contour of the outside of the loop is the opposite direction, one contour point is determined for each line. Rasterization, incrementing the value of each pixel on a predetermined one direction side divided by the contour line for each line depending on whether the contour line increases or decreases with respect to the line processing direction and the vertical direction, or A figure processing method characterized in that decrementing is performed. 3. A graphic processing method used in the graphic processing device according to claim 1, wherein 1-pixel data is expressed by a signed integer of a plurality of bits or an unsigned integer with an offset, and 1 data of the filling means. A graphic processing method characterized by storing data for a plurality of pixels in a word and performing increment or decrement processing in units of one data word or an integral multiple thereof. 4. The graphic processing method according to claim 3, wherein a plurality of bit portions corresponding to a direction opposite to the line processing direction are all cleared from a plurality of bit portions corresponding to the filling start point and correspond to the filling start point. Increment operation by adding the added value that sets each least significant bit of the multi-bit part and the multi-bit part corresponding to the forward direction of the line processing direction from the multi-bit part and clears the other bits If the multi-bit part corresponding to the direction opposite to the line processing direction is all set from the multi-bit part corresponding to the filling start point and the multi-bit part corresponding to the filling start point is not at the lowest position of the data word, Multiple bits corresponding to the filling start point, with the least significant bit cleared and all other bits set If the multi-bit part corresponding to the forward direction of the line processing direction is not at the bottom of the data word, the least significant bit is cleared and each other bit is set, and the least-significant multi-bit part is all set. A figure processing method, characterized in that a decrement operation is performed by adding the added value obtained in step (1).