CN107220224A - A kind of literary generation method of square seedling derived based on intelligence - Google Patents

Abstract

The present invention relates to a text generation method, in particular to a block Miao text generation method based on intelligent derivation; the method includes the following steps: A: establishing component bounding boxes and normalizing them; B: determining the bounding box topology Transformation coefficient; C: transform the component bounding box; D: establish the coding tree corresponding to the derivation process; E: establish the intelligent derivation model; F: generate the block Miaowen according to the intelligent derivation algorithm. The square Hmong characters generated by the invention are neat and standardized, do not occupy coding space, and can meet practical requirements.

Description

A method for generating block Miao characters based on intelligent derivation

technical field

The invention relates to a method for generating characters, in particular to a method for generating square Hmong characters based on intelligent derivation.

Background technique

When current operating systems process ideographic characters, they usually encode and design characters "word by word". The storage of character data is huge, and it is difficult to reduce the size of the coded character set. Square Miao script is the main carrier of folk Miao culture in the Wuling Mountain area. It is an ideographic script with fixed structural characteristics, and it is basically a combination of characters. Because of its late start in information processing research, Unicode and other character set standards have not assigned code points to it. In the prior art, there is a method for dynamically generating block Miao characters based on component combination operation expressions. By using the OpenType font technology, constructing the font model of the square Miao script, planning the font layout table, and writing the mapping script for combination and replacement, the Hmong characters of different structures can be generated without occupying the Unicode coding interval. However, this method must be based on the character data of the square Hmong script, and it is necessary to make a font for each square Hmong script. The block Hmong characters without fonts will not be displayed.

Contents of the invention

The purpose of the present invention is to provide a method for generating square Miao characters based on intelligent derivation, which solves the problem that the storage of character data is huge and the coded character set is difficult to slim down.

The present invention is achieved like this, and a kind of square Miaowen generation method based on intelligent derivation comprises the following steps:

A: Establish component bounding boxes and normalize them;

B: Determine the bounding box topology transformation coefficient;

C: Transform the bounding box of the component;

D: Establish a coding tree corresponding to the derivation process;

E: Establish an intelligent derivation model;

F: Generate square Miao characters according to the intelligent derivation algorithm.

A further technical solution of the present invention is: said step A is: given a Hmong character grid, its width and height are respectively W and H; with the upper left corner of the grid as the origin, establish an X-axis to the right and a Y-axis Downward Cartesian coordinate system; in this coordinate system, the component bounding box is a minimum circumscribed rectangle determined by the coordinates (x, y) of the upper left vertex of the component character image and the width (w) and height (h), which can represent is a quadruple B: B=(x, y, w, h). Among them, x ∈ (0, W), y ∈ (0, H), w ∈ (0, Wx), h ∈ (0, Hy); and the component bounding box is normalized as:

A further technical solution of the present invention is: said step B is: Given a bounding box B=(x, y, w, h) and B'=(x', y', w', h'), transform B The topological transformation coefficient of B' is calculated by formula (1):

A further technical solution of the present invention is: in the step C, calculate the position and size of the target component in the Hmong text, involving centering, tiling and aspect ratio adjustment transformation operations;

in the center

After adjustment, the width and height of the bounding box remain unchanged, and only the coordinates of the upper left vertex change. Given bounding box B ₁ =(x ₁ ,y ₁ ,w ₁ ,h ₁ ) and B ₂ =(x ₂ ,y ₂ ,w ₂ ,h ₂ ), place B ₁ along the X-axis direction, Y-axis direction, At the same time, it is centered in B ₂ along the X-axis and Y-axis direction. In the three cases, the x' and y' in the new bounding box B' ₁ = (x', y', w ₁ , h ₁ ) are respectively used in the formula ( 2), (3), (4) calculation:

tiling

The above bounding box B ₁ is tiled into the bounding box B ₂ along the X-axis direction and the Y-axis direction, and the new bounding box B'=(x ₂ ,y ₁ ,w ₂ , h ₁ ), B'=(x ₁ , y ₂ , w ₁ , h ₂ ), and B'=(x ₂ , y ₂ , w ₂ , h ₂ ).

aspect ratio adjustment

Set the bounding box of the original component as B=(x, y, w, h), and the bounding box obtained after adjusting the aspect ratio to r is B'; then, for the left and right components, keep the height unchanged, according to the formula (5 ) to adjust; for upper and lower components, keep the width constant, and adjust according to formula (6); for outsourcing components or internal components, both height and width are changed, if h/w≤r, then follow formula (5) Adjust, otherwise, adjust according to formula (6).

B'=(x,y,h/r,h) (5)

B'=(x,y,w,w.r) (6)

The further technical scheme of the present invention is: described step D is: utilize the recursion of binary tree, the solution of each node data in the tree is decomposed into simple subtree to solve, establishes the coding tree of corresponding derivation process; Coding tree is derivation Symbolic description of the tree; in the coding tree, the seven Hmong structures of the left-right type, the up-down type, the inner-outer type, and the upper-left, lower-left, and upper-right structures of the side-wrapping type are respectively represented by the letters "Z" and "S". , "N", "A", "B" and "C"; the components are represented by the Unicode code of their standard characters; The coding tree of the text is a 2-level structure, divided into 3 layers.

The further technical scheme of the present invention is: described step E is to set up the Miao Wen intelligent derivation model by 3 levels; The top layer is the submodule that completes the coding tree decomposition work; It is used to solve the topological transformation coefficient of each node in the coding tree; the bottom layer is a sub-module for adjusting the aspect ratio of components; since the left and right structures and the upper and lower structures can be classified as vertical and horizontal structures, the inner and outer structures and the three side structures It is classified as the inclusion structure, so the middle layer only designs two structure derivation sub-modules for the vertical and horizontal structure and the inclusion structure.

The further technical scheme of the present invention is: described step F comprises following algorithm:

Algorithms for top-level submodules:

1) According to the Unicode coding sequence of the components contained in the Hmong script and the Hmong script structure type, a corresponding Miao script generation coding tree is established;

2) carry out depth-first root-first traversal to the coding tree, and obtain the generated coding sequence of Hmong;

3) Analyze and generate the coding sequence, call the corresponding structure derivation sub-module according to the process, and deduce the transformation parameters of all the components contained in the Miao language;

4) According to the obtained component transformation parameters, transform the original component into the target component in Miao text. Algorithm of the vertical and horizontal structure derivation sub-module:

1) Determine its proportion in the entire Hmong structure according to the standard bounding box of each component, and determine its maximum possible bounding box according to the order in which it appears in the Hmong structure;

2) For the left and right structures, center each component along the X-axis direction to the largest bounding box of the Miao structure according to formula (2); for the upper and lower structures, center each component along the Y-axis direction to the maximum bounding box of the Miao structure In the largest bounding box;

3) calling the aspect ratio adjustment sub-module to adjust the aspect ratio of each component to obtain the bounding box of each component in Hmong;

4) Use formula (1) to calculate the conversion coefficient of each component.

Derivation algorithms with structure derivation submodules:

1) Take the largest bounding box of the outsourcing component as its actual bounding box in Hmong;

2) Call the aspect ratio adjustment submodule to adjust the standard bounding box containing components to the size of the largest bounding box in Hmong;

3) tile the adjusted bounding box of the inner component into the largest bounding box of the outer component along the X-axis and Y-axis direction at the same time, and obtain its actual bounding box in Hmong;

4) Use formula (1) to calculate the transformation coefficient of each component.

Algorithm of the aspect ratio adjustment submodule:

1) For components with vertical and horizontal structures, check whether the aspect ratio is between the minimum value and the maximum value. If so, then keep the current aspect ratio constant; otherwise, adjust the aspect ratio of the left and right type components to the average value according to formula (5), and adjust the aspect ratio of the up and down type components to the average value according to formula (6);

2) For components containing structures, compare their aspect ratios to the average. If it is equal, keep the current aspect ratio unchanged; if it is greater than, adjust the aspect ratio to the average value according to formula (6); if it is less than, then adjust the aspect ratio to the average value according to formula (5).

The beneficial effects of the present invention are: the square Miao characters generated based on the target component font data obtained by intelligent derivation are neat and standardized, and do not occupy coding space, and can meet the practical requirements; the method has three characteristics: first, based on The idea of generating square Miao characters by component transformation is the core of the model; secondly, using the recursion of the binary tree to realize the solution of component transformation parameters not only ensures the simplicity and effectiveness of the method of generating square Miao characters, but also embodies the advantages of character derivation. Intelligence; thirdly, in this model, all kinds of data required for intelligent derivation generated by Cube Miaowen are stored in an independent information database, so that the language features of Cube Miaowen can be separated from the program logic of word processing software, which is convenient Adjustment, optimization and expansion of software functions will be carried out in the future.

Description of drawings

Fig. 1 is a kind of block Miao script structure schematic diagram provided by the present invention;

Fig. 2 is a kind of DFA state diagram of identifying square Miaowen provided by the present invention;

Fig. 3 is a schematic diagram of a component bounding box provided by the present invention;

Fig. 4 is a schematic structural diagram of an intelligent derivation system provided by the present invention;

Fig. 5 is the derivation flow chart of a kind of block Miao character generation method based on intelligent derivation provided by the present invention;

Fig. 6 is a derivation schematic diagram of a block Hmong script example C ₆ provided by the present invention;

Fig. 7 is a derivation result of a square Hmong script example C ₆ provided by the present invention.

detailed description

Embodiment one:

Figures 1-7 are schematic diagrams of a method for generating block Miao characters based on intelligent derivation. The structure of square Miaowen can be roughly divided into 4 types as shown in Figure 1: left and right type, up and down type, side enclosure type and internal and external type. Among them, the side circumference type can be divided into three types: left upper bag, left lower bag and right upper bag. According to the character-making principle and character-building method of square Miao characters, when a Miao character consists of 3 or more parts, some two or 3 parts can usually form a simple Chinese character. At this time, it is advisable to regard this simple Chinese character as a component. The statistical results of the 1129 square Miao scripts that have been sorted out show that most of the Miao scripts can be regarded as two-component type, and only a few left-right and up-down type Miao scripts need to be treated as three-component type.

The components of square Miao script can be classified into 16 categories: left component (c _l ), right component (c _r ), upper component (c _u ), lower component (c _d ), upper left outer component (c _olu ), lower right inner component Containing components (c _ird ), lower left outsourcing components (c _old ), upper right containing components ( _ciru ), upper right outsourcing components (c _oru ), lower left containing components ( _cild ), fully outsourcing components (c _oa ), all Contains component (c _ia ), right left component (c _rl ), right right component (c _rr ), lower left component (c _dl ) and lower right component (c _dr ). If represent square Miao characters with SG, C _i represents component character (i=start, end, first, middle, final), then the composition of square Miao characters can be represented as SG={C _start C _end } or SG={ C _first C _middle C _final }. Among them, C _start = {c _l , c _u , c _olu , c _old , c _oru , c _oa }, C _end = {c _r , c _d , c _ird , c _iru , c _ild , c _ia }, C _first = {c _l , c _u }, C _middle = {c _rl , c _dl }, C _final = {c _rr , c _dr }.

A square Miao script consists of 2-3 components. Based on the 16 types of components, a Deterministic Finite Automata (DFA) for recognizing block Miao characters can be constructed. The DFA is expressed as M=(K, A, F, S, Z) with a quintuple, where:

1) K is a finite set of states, K={0, 1, 2, 3, 4, 5, 6, 7, 8, 9};

2) A is a finite set of input symbols, A={c _l , c _u , c _olu , c _old , c _oru , c _oa , c _r , c _d , c _ird , c _iru , c _ild , c _ia , c _rl , c _dl , c _rr , c _dr };

3) F is a mapping set on K×A→K, F={f(0,c _l )=1, f(0,c _u )=2, f(0,c _olu )=3, f(0 ,c _old )=4, f(0,c _oru )=5, f(0,c _oa )=6, f(1,c _rl )=7, f(1,c _r )=9, f(2 ,c _dl )=8, f(2,c _u )=9, f(3, _cird )=9, f(4, _ciru )=9, f(5, _cild )=9, f(6 ,c _ia )=9, f(7,c _rr )=9, f(8,c _dr )=9};

4) S∈K is the initial state with unique value, S=0;

5) is the final state set (the final state is also called end state, acceptable state or identifiable state), Z={9}.

Figure 2 shows the state diagram representation of the DFA. Starting from the initial state 0 in Figure 2, the component character sequences on any path to the final state 9 (indicated by double circles) are synthesized into a square Hmong script.

In this paper, the Chinese characters, Chinese radicals and pure symbol characters used as square Hmong script components are called original components, and the components displayed in square Hmong script are called target components. The target component is obtained from the original component through topology transformation. When all the target components contained in a Hmong script have been transformed from the original components, the square Miao script can be directly combined from these target components.

A bounding box is a simple and geometric figure that can correctly reflect the position and size characteristics of complex geometric objects. In order to simplify the component transformation operation, the processing of Hmong component character images in this paper is transformed into the processing of its bounding boxes. For this reason, the definition of the bounding box of the Miao component is given as follows:

Definition 1 Given a square grid of Hmong characters, its width and height are W and H respectively. With the upper left corner of the grid as the origin, establish a Cartesian coordinate system with the X axis right and the Y axis down. In this coordinate system, the component bounding box is a minimum circumscribed rectangle determined by the upper left vertex coordinates (x, y) and the width (w) and height (h) of the component character image (as shown in Figure 2), which can represent is a quadruple B: B=(x,y,w,h). Among them, x ∈ (0, W), y ∈ (0, H), w ∈ (0, Wx), h ∈ (0, Hy); for the convenience of analysis, according to the principle of topology invariance, the component bounding box is normalized processed as: In this paper, the bounding box refers to the normalized bounding box.

The generation process of square Miaowen can be transformed into the calculation process of component bounding box. This process is the process of gradually adjusting the standard bounding box of the original component to the bounding box of the target component by calculating and determining the topological transformation coefficient.

Given a bounding box B = (x, y, w, h) and B' = (x', y', w', h'), the topological transformation coefficient of transforming B into B' is calculated by formula (1):

Since the standard bounding box information of the original component can be obtained directly from the corresponding Chinese character font, the problem of determining the topological transformation coefficient is transformed into the problem of solving the bounding box of the target component.

Solving the bounding box of the target component is to calculate the position and size of the target component in Miao text, which involves transformation operations such as centering, tiling, and aspect ratio adjustment.

1) centered

Centering refers to adjusting one bounding box to the middle position of another bounding box along a certain coordinate axis. After adjustment, the width and height of the bounding box remain unchanged, and only the coordinates of the upper left vertex change. Given bounding box B ₁ =(x ₁ ,y ₁ ,w ₁ ,h ₁ ) and B ₂ =(x ₂ ,y ₂ ,w ₂ ,h ₂ ), place B ₁ along the X-axis direction, Y-axis direction, At the same time, it is centered in B ₂ along the X-axis and Y-axis direction, and the x' and y' in the new bounding box B' ₁ = (x', y', w ₁ , h ₁ ) are obtained in the three cases respectively using the formula ( 2), (3), (4) calculation:

2) Tile

Tiling refers to expanding or compressing a bounding box into another bounding box along a certain axis. The above bounding box B ₁ is tiled into the bounding box B ₂ along the X-axis direction and the Y-axis direction, and the new bounding box B'=(x ₂ ,y ₁ ,w ₂ , h ₁ ), B'=(x ₁ , y ₂ , w ₁ , h ₂ ), and B'=(x ₂ , y ₂ , w ₂ , h ₂ ).

3) Aspect ratio adjustment

Aspect ratio is an important parameter to reflect the shape of components in Hmong. Different types of components have different methods for aspect ratio adjustment. Let the bounding box of the original component be B=(x, y, w, h), and the bounding box obtained after adjustment according to the aspect ratio r is B'. Then, for the left and right components, keep the height constant, and adjust according to formula (5); for the up and down components, keep the width constant, and adjust according to formula (6); Change, if h/w≤r, adjust according to formula (5), otherwise, adjust according to formula (6).

B'=(x,y,h/r,h) (5)

B'=(x,y,w,w.r) (6)

The key to solving the bounding box of the target component is to obtain the topological transformation coefficient and the position and size data of the bounding box. In this paper, the binary tree in the form of "structure + component" is used to represent the derivation process of Miaowen generation. The non-leaf nodes of the derivation tree represent the Miao structure, and the leaf nodes represent specific components. By using the recursion of the binary tree, the solution of each node data in the tree can be decomposed into simple subtrees for solution. A coding tree is a symbolic description of a derivation tree. In the coding tree, the seven Hmong structures of the left-right type, the up-down type, the inner-outer type, and the upper-left, lower-left, and upper-right structures of the side-wrapping type are respectively represented by the letters "Z", "S", "N", "A","B" and "C"represent; the components are represented by the Unicode encoding of their standard characters. The coding tree of the 2-component Miao language is a first-level structure, divided into two layers. The coding tree of the 3-component Miao language is a 2-level structure, divided into 3 layers. The codes of C ₆ and C ₈ in Figure 1 are "S 5408Z76EE 76EE" and "B 51A0 5B50" respectively.

The Miao Wen intelligent derivation model can be structured according to the three levels shown in Figure 4. The top layer is the sub-module that completes the coding tree decomposition work; the middle layer is the sub-module that deduces the structure for different structures of Hmong texts, and is used to solve the topological transformation coefficient of each node in the coding tree. The bottom layer is a sub-module for component aspect ratio adjustment. Considering the similarity of the derivation process, the left-right structure and the up-down structure can be classified as vertical and horizontal structures, and the internal and external structures and the three side structures can be classified as inclusion structures. Therefore, the middle layer only needs to design two structure derivation sub-modules for the vertical and horizontal structure and the containment structure.

The top-level submodule first traverses the entire coding tree in depth-first order, and decomposes the coding tree into multiple subtrees in turn. Every time a subtree is decomposed, the subtree is handed over to the corresponding structure derivation submodule for solution, and the solution result is used to replace the subtree. This loop continues until all subtrees are solved.

The algorithm design of the top-level sub-module is as follows:

1) According to the Unicode coding sequence of the components contained in the Hmong script and the structure type of the Hmong script, the corresponding Hmong script generated coding tree is established.

2) Perform depth-first root-first traversal on the coding tree to obtain the generated coding sequence of Hmong.

3) Analyze and generate the coding sequence, call the corresponding structure derivation sub-module according to the process shown in Figure 5, and deduce the transformation parameters of all the components contained in the Hmong script.

4) According to the obtained component transformation parameters, transform the original component into the target component in Miao text.

The components of the vertical and horizontal structures in Miao scripts, the shapes that appear in Miao scripts are not only related to the order in which they appear in Miao scripts, but also related to their size. The proportion of the component is proportional to the size of the original component. The bounding box of the 2-component Hmong component is directly taken from the standard bounding box of the original component. The bounding box of the 3-component Hmong component is the result of depth-first traversal of the spanning tree. The relevant information of the bounding box of each component is stored in the component position information database. The library specifically contains data such as component name, component type, structure name, number of positions, and list of maximum bounding boxes for each position. When solving, directly take relevant data from the library and act on each component to obtain the maximum bounding box of the corresponding component.

The algorithm design of the vertical and horizontal structure derivation sub-module is as follows:

1) According to the standard bounding box of each component, determine its proportion in the entire Hmong structure, and determine its maximum possible bounding box according to the order in which it appears in the Hmong structure.

2) For the left and right structures, center each component along the X-axis direction to the largest bounding box of the Miao structure according to formula (2); for the upper and lower structures, center each component along the Y-axis direction to the maximum bounding box of the Miao structure in the largest bounding box.

3) Call the aspect ratio adjustment sub-module to adjust the aspect ratio of each component to obtain the bounding box of each component in Hmong.

4) Use formula (1) to calculate the conversion coefficient of each component.

The common feature of contained structure Miaowen is that the outsourcing component acts as a frame, the inner component is embedded in this frame, and the largest possible bounding box of each component is determined by the outsourcing component. The relevant information of the bounding box of each component is also stored in the component location information database. When solving, directly take relevant data from the library and act on each component to obtain the maximum bounding box of the corresponding component.

The derivation algorithm including the structure derivation sub-module is designed as follows:

1) Take the largest bounding box of the outsourcing component as its actual bounding box in Hmong.

2) Call the aspect ratio adjustment submodule to adjust the standard bounding box containing the components to the size of the largest bounding box in Hmong.

3) Flatten the adjusted bounding box of the inner component into the largest bounding box of the outer component along the X-axis and Y-axis to obtain its actual bounding box in Hmong.

4) Use formula (1) to calculate the transformation coefficient of each component.

During intelligent derivation in Miao language, each structure derivation sub-module needs to call the aspect ratio adjustment sub-module, and adjust the components according to different adjustment methods according to the relevant data read from the component aspect ratio information database. The component aspect ratio information database contains component name, component type, structure name, component occurrence times, and the minimum value, maximum value, average value, variance and other data of the aspect ratio.

The algorithm design of the aspect ratio adjustment sub-module is as follows:

1) For components with vertical and horizontal structures, check whether the aspect ratio is between the minimum value and the maximum value. If so, keep the current aspect ratio unchanged; otherwise, adjust the aspect ratio of the left and right components to the average value according to formula (5), and adjust the aspect ratio of the up and down components to the average value according to formula (6).

2) For components containing structures, compare their aspect ratios to the average. If it is equal, keep the current aspect ratio unchanged; if it is greater than, adjust the aspect ratio to the average value according to formula (6); if it is less than, then adjust the aspect ratio to the average value according to formula (5).

Based on the above algorithm, the generation and solving process of Miao characters can be decomposed into a series of character generation and solving of the first-level structure starting from the coding leaf node. The following uses Miaowen _C6 as an example to illustrate the intelligent derivation and solution process generated by Miaowen. According to Figure 6, the steps to solve the component transformation coefficient of Miaowen _C6 are described as follows:

1) Determine the structural progression. C ₆ has 2 levels of structure: "S" for the top level structure and "Z" for the last level structure.

2) Separate the final structure and solve it as a pseudo-Miao text, as shown in Figure 7(a). The generated code sequence at this time is "Z 76EE 76EE". Use the vertical and horizontal structure to deduce the sub-modules, and find the bounding boxes of the two sub-components "76EE" and "76EE" in the pseudo-Miao script.

3) Treat the pseudo-Miao language (represented as "#1") as a component and incorporate it into the upper-level structure, as shown in Figure 7(b). At this time, the generated coding sequence is "S 5408#1". Using the vertical and horizontal structure derivation sub-module, find the two components "5408" and "#1" as the bounding box of the target component in C ₆ .

4) According to formula (1), calculate the topological transformation coefficient of each component in Miao script.

5) Transform each component according to the above-mentioned topological transformation coefficient to obtain the corresponding target component, and combine these targets together to obtain the Hmong character C ₆ .

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention should be included in the protection of the present invention. within range.

Claims (7)

1. A method for generating block Miao characters based on intelligent derivation, characterized in that, the method may further comprise the steps: A: Establish component bounding boxes and normalize them; B: Determine the bounding box topology transformation coefficient; C: Transform the bounding box of the component; D: Establish a coding tree corresponding to the derivation process; E: Establish an intelligent derivation model; F: Generate square Miao characters according to the intelligent derivation algorithm. 2. A kind of square Miao character generation method based on intelligent derivation according to claim 1, is characterized in that: described step A is: given a square Hmong character grid, its width and height are respectively W and H ;Taking the upper left corner of the grid as the origin, establish a Cartesian coordinate system with the X axis right and the Y axis downward; A minimum circumscribed rectangle determined jointly by (w) and height (h) can be expressed as a quadruple B: B=(x, y, w, h); where, x∈(0,W), y∈( 0,H),w∈(0,Wx),h∈(0,Hy); and normalize the component bounding box as: 3. A kind of square Miaowen generation method based on intelligent derivation according to claim 2, is characterized in that: described step B is: given bounding box B=(x, y, w, h) and B'= (x', y', w', h'), the topological transformation coefficient of transforming B into B' is calculated by formula (1): <mrow> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <msub> <mi>R</mi> <mi>w</mi> </msub> <mo>=</mo> <msup> <mi>w</mi> <mo>&amp;prime;</mo> </msup> <mo>/</mo> <mi>w</mi> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msub> <mi>R</mi> <mi>h</mi> </msub> <mo>=</mo> <msup> <mi>h</mi> <mo>&amp;prime;</mo> </msup> <mo>/</mo> <mi>h</mi> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msup> <mi>x</mi> <mo>,</mo> </msup> <mo>=</mo> <mi>x</mi> <mo>+</mo> <mi>x</mi> <mo>.</mo> <msub> <mi>R</mi> <mi>w</mi> </msub> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msup> <mi>y</mi> <mo>,</mo> </msup> <mo>=</mo> <mi>y</mi> <mo>+</mo> <mi>y</mi> <mo>.</mo> <msub> <mi>R</mi> <mi>h</mi> </msub> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>.</mo> </mrow> 4. A method for generating block Miao characters based on intelligent derivation according to claim 3, characterized in that: in the step C, calculating the position and size of the target component in the Hmong characters involves centering and tiling and aspect ratio adjustment transform operations; in the center After adjustment, the width and height of the bounding box remain unchanged, and only the coordinates of the upper left vertex change; the given bounding box B ₁ =(x ₁ ,y ₁ ,w ₁ ,h ₁ ) and B ₂ =(x ₂ ,y ₂ ,w ₂ , h ₂ ), center B ₁ into B ₂ along the X-axis direction, the Y-axis direction, and along the X-axis and Y-axis directions at the same time, and obtain a new bounding box B' ₁ =(x',y' in three cases ,w ₁ ,h ₁ ) x' and y' are calculated by formulas (2), (3) and (4) respectively: <mrow> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <msup> <mi>x</mi> <mo>,</mo> </msup> <mo>=</mo> <msub> <mi>x</mi> <mn>2</mn> </msub> <mo>+</mo> <mo>|</mo> <msub> <mi>w</mi> <mn>1</mn> </msub> <mo>-</mo> <msub> <mi>w</mi> <mn>2</mn> </msub> <mo>|</mo> <mo>/</mo> <mn>2</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msup> <mi>y</mi> <mo>,</mo> </msup> <mo>=</mo> <msub> <mi>y</mi> <mn>1</mn> </msub> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow> <mrow> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <msup> <mi>x</mi> <mo>,</mo> </msup> <mo>=</mo> <msub> <mi>x</mi> <mn>1</mn> </msub> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msup> <mi>y</mi> <mo>,</mo> </msup> <mo>=</mo> <msub> <mi>y</mi> <mn>2</mn> </msub> <mo>+</mo> <mo>|</mo> <msub> <mi>h</mi> <mn>1</mn> </msub> <mo>-</mo> <msub> <mi>h</mi> <mn>2</mn> </msub> <mo>|</mo> <mo>/</mo> <mn>2</mn> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>3</mn> <mo>)</mo> </mrow> </mrow> <mrow> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <msup> <mi>x</mi> <mo>,</mo> </msup> <mo>=</mo> <msub> <mi>x</mi> <mn>2</mn> </msub> <mo>+</mo> <mo>|</mo> <msub> <mi>w</mi> <mn>1</mn> </msub> <mo>-</mo> <msub> <mi>w</mi> <mn>2</mn> </msub> <mo>|</mo> <mo>/</mo> <mn>2</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msup> <mi>y</mi> <mo>,</mo> </msup> <mo>=</mo> <msub> <mi>y</mi> <mn>2</mn> </msub> <mo>+</mo> <mo>|</mo> <msub> <mi>h</mi> <mn>1</mn> </msub> <mo>-</mo> <msub> <mi>h</mi> <mn>2</mn> </msub> <mo>|</mo> <mo>/</mo> <mn>2</mn> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </mrow> tiling The above bounding box B ₁ is tiled into the bounding box B ₂ along the X-axis direction and the Y-axis direction, and the new bounding box B'=(x ₂ ,y ₁ ,w ₂ , h ₁ ), B'=(x ₁ ,y ₂ ,w ₁ ,h ₂ ) and B'=(x ₂ ,y ₂ ,w ₂ ,h ₂ ); aspect ratio adjustment Set the bounding box of the original component as B=(x, y, w, h), and the bounding box obtained after adjusting the aspect ratio to r is B'; then, for the left and right components, keep the height unchanged, according to the formula (5 ) to adjust; for upper and lower components, keep the width constant, and adjust according to formula (6); for outsourcing components or internal components, both height and width are changed, if h/w≤r, then follow formula (5) Adjust, otherwise, adjust according to formula (6); B'=(x,y,h/r,h) (5) B'=(x,y,w,w.r) (6). 5. a kind of square Miao Wen generation method based on intelligent derivation according to claim 4, it is characterized in that: described step D is: utilize the recursion of binary tree, the solution of each node data in the tree is decomposed into simple Solve in the subtree, and establish a coding tree corresponding to the derivation process; the coding tree is a symbolic description of the derivation tree; The seven Hmong structures are represented by the letters "Z", "S", "N", "A", "B" and "C" respectively; the components are represented by the Unicode codes of their standard characters; 2. The coding tree of the text has a 1-level structure, which is divided into 2 layers; the coding tree of the 3-component Miao language is a 2-level structure, which is divided into 3 layers. 6. a kind of square Hiao Wen generation method based on intelligent derivation according to claim 5, is characterized in that: described step E is: set up Miao Wen intelligent derivation model by 3 levels; sub-module; the middle layer is a sub-module for structure derivation for different structures of Miao texts, which is used to solve the topological transformation coefficient of each node in the coding tree; the bottom layer is a sub-module for adjusting the aspect ratio of components; since the left and right structures and The upper and lower structures are classified as vertical and horizontal structures, and the internal and external structures and the three side structures are classified as inclusion structures. Therefore, the middle layer only designs two structure derivation sub-modules for vertical and horizontal structures and inclusion structures. 7. a kind of square Miao language generation method based on intelligent derivation according to claim 6, is characterized in that: described step F comprises following algorithm: Algorithms for top-level submodules: 1) According to the Unicode coding sequence of the components contained in the Hmong script and the Hmong script structure type, a corresponding Miao script generation coding tree is established; 2) carry out depth-first root-first traversal to the coding tree, and obtain the generated coding sequence of Hmong; 3) Analyze and generate the coding sequence, call the corresponding structure derivation sub-module according to the process, and deduce the transformation parameters of all the components contained in the Miao language; 4) Transform the original component into the target component in Miao text according to the obtained component transformation parameters; Algorithm of the vertical and horizontal structure derivation sub-module: 1) Determine its proportion in the entire Hmong structure according to the standard bounding box of each component, and determine its maximum possible bounding box according to the order in which it appears in the Hmong structure; 2) For the left and right structures, center each component along the X-axis direction to the largest bounding box of the Miao structure according to formula (2); for the upper and lower structures, center each component along the Y-axis direction to the maximum bounding box of the Miao structure In the largest bounding box; 3) calling the aspect ratio adjustment sub-module to adjust the aspect ratio of each component to obtain the bounding box of each component in Hmong; 4) Use formula (1) to calculate the conversion coefficient of each component; Derivation algorithms with structure derivation submodules: 1) Take the largest bounding box of the outsourcing component as its actual bounding box in Hmong; 2) Call the aspect ratio adjustment submodule to adjust the standard bounding box containing components to the size of the largest bounding box in Hmong; 3) tile the adjusted bounding box of the inner component into the largest bounding box of the outer component along the X-axis and Y-axis direction at the same time, and obtain its actual bounding box in Hmong; 4) Use formula (1) to calculate the conversion coefficient of each component; Algorithm of the aspect ratio adjustment submodule: 1) For components with vertical and horizontal structures, check whether the aspect ratio is between the minimum value and the maximum value; if so, keep the current aspect ratio unchanged; otherwise, adjust the aspect ratio of the left and right components to the average value according to formula (5) , adjust the aspect ratio of the upper and lower members to the average value according to formula (6); 2) For the component containing the structure, compare its aspect ratio with the average value; if it is equal, keep the current aspect ratio unchanged; if it is greater than, adjust the aspect ratio to the average value according to formula (6); if it is less than, then Adjust the aspect ratio to the average value according to formula (5).