Method for generating tetragonal continuous chartlet from any picture

Abstract

The invention discloses a method for generating a tetragonal continuous chartlet from any picture, and the tetragonal continuous chartlet defines four texture coordinate sets for each vertex, so that textures are continuously laid on the surface of a patch. The four texture coordinates respectively correspond to four corners of the patch so that the texture can be correctly mapped around each vertex. Then, smooth interpolation is carried out on the texture coordinates in the surface patch by using an interpolation technology, so that continuous texture mapping is realized on the whole surface patch; therefore, the problems of texture distortion and stretching can be reduced, and the sense of reality of surface details is improved. The method provided by the invention solves the problems of texture distortion and unsmooth edge transition when the continuous chartlet is generated in the traditional technology, and the complexity and the calculation cost of the method are better than those of the traditional method.

Description

Technical Field

[0001] The invention relates to a method for generating a four-sided continuous map from any picture, and belongs to the field of computers. Background Art

[0002] Four-Directional Continuous Texture Mapping is a technique used for texture mapping in computer graphics. It is often used to enhance the appearance of three-dimensional models and make their surfaces look more realistic.

[0003] In traditional texture mapping, each triangle patch has only one set of texture coordinates to determine its sampling position in the texture image. However, when the surface curvature of the triangle is high, this mapping method will cause the texture map to stretch or distort on the surface, thus affecting the visual effect.

[0004] Quad continuous mapping allows the texture to be tiled continuously on the surface of the patch by defining four sets of texture coordinates for each vertex. These four texture coordinates correspond to the four corners of the patch, so that the texture can be correctly mapped around each vertex. Then, the texture coordinates are smoothly interpolated inside the patch using interpolation techniques, so that continuous texture mapping is achieved across the entire patch. This can reduce the problem of texture distortion and stretching, and improve the realism of surface details.

[0005] Quad maps are often used to model and render curved surfaces, such as terrain, characters, etc., to enhance their appearance and provide more realistic texture effects.

[0006] Problems and shortcomings of existing technologies:

[0007] 1. Texture distortion: In a four-dimensional continuous map, the interpolation of texture coordinates may cause the texture to stretch, twist or distort. Especially when there are areas with height changes or large curvature on the surface of the patch, the texture mapping result may be inaccurate, resulting in visual discontinuity or unnaturalness.

[0008] 2. Edge transition problem: Between adjacent patches, due to the interpolation of texture coordinates, the edges of the texture may have discontinuous or non-smooth transitions. This will cause visual gaps or breaks between patches, destroying the continuity of the overall texture.

[0009] 3. Requires higher texture resolution: In order to maintain the details and accuracy of the texture, quad-continuous mapping usually requires higher texture resolution. This will increase storage and computing costs, and impose additional burdens on hardware resources and performance.

[0010] 4. Complexity and computational cost: Implementing four-dimensional continuous mapping requires interpolation calculation of texture coordinates. Especially in large-scale scenes or complex models, the amount of calculation may be large, resulting in reduced rendering performance.

[0011] 5. Not applicable to non-quadrilateral patches: The method of quadrilateral continuous mapping is not applicable to non-quadrilateral patches. For patches with more complex shapes, such as polygons or irregular surfaces, it may be more difficult to achieve continuous texture mapping.

[0012] Although quad-continuous mapping can provide better texture mapping effects in some cases, it also has these disadvantages, which need to be weighed and improved in practical applications. Summary of the invention

[0013] In order to solve the above technical problems, a method for generating a four-square continuous map from any image is provided, which solves the problems of texture distortion and uneven edge transition when generating continuous maps in traditional technologies, and the complexity and computational cost of this method are better than those of traditional methods. It is specifically implemented through the following technical solutions:

[0014] A method for generating a four-square continuous map from an arbitrary image, including a cycle block, a seed block and a candidate block;

[0015] The seed block is used as an independent image block;

[0016] The cycle block is composed of a plurality of seed blocks and is used for seamless tiling in the map;

[0017] The candidate blocks are candidate image blocks for replacing the seed blocks. In the process of generating the four-way continuous map, according to certain algorithms and rules, the system will select a suitable replacement block from a set of predefined candidate blocks to replace the seed block.

[0018] Furthermore, the cycle block is formed by arranging and connecting the seed blocks according to a certain rule.

[0019] Furthermore, the edges of adjacent seed blocks are seamlessly connected to form a smooth transition effect during the tiling process.

[0020] Furthermore, the method for generating a four-sided continuous map from an arbitrary image includes the steps of: determining a loop region of appropriate size based on an input image; determining a seed block of the image according to the size of the loop region; extracting the best matching candidate block according to the seed block; continuing to fill according to the candidate block until the entire loop block is filled; and finally splicing all the loop blocks to ensure the best match at the splicing point to generate a final image.

[0021] Furthermore, the method for generating a four-dimensional continuous map from any picture comprises the steps of:

[0022] Determine the size of the loop region based on the input image, cut the original image into small blocks, and ensure that each small block can be tiled seamlessly; determine the seed block based on the loop region, and a good seed block should have good image continuity, that is, its edges can be seamlessly connected to form a visually smooth effect;

[0023] Determine candidate blocks according to the seed blocks, and set a transition area of ​​appropriate size at the edge when splicing the candidate blocks;

[0024] The transition area usually has obvious discontinuous parts. Compare the similarity of the two blocks, find the transition line with the largest difference, and use the Gaussian blur method of pixels to optimize the transition area.

[0025] Continue to fill in candidate blocks until the entire loop area is filled; after completing the above steps, if there are still discontinuous areas, you can set a frame selection area for secondary optimization.

[0026] The present invention also provides any of the above methods for generating a four-dimensional continuous map from any image for computer texture mapping.

[0027] Compared with the prior art, the present invention has the following advantages: 1. The present invention ensures the continuity of the generated map and avoids the sense of repetition through the optimal matching between the seed block, the candidate block and the cycle block; 2. The speed is fast and the accuracy is high for general texture maps. DETAILED DESCRIPTION

[0028] The technical solution of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] A method for generating a four-square continuous map from an arbitrary image, including a cycle block, a seed block and a candidate block;

[0030] The seed block is used as an independent image block;

[0031] The cycle block is composed of a plurality of seed blocks and is used for seamless tiling in the map;

[0032] The candidate blocks are candidate image blocks for replacing the seed blocks. In the process of generating the four-way continuous map, according to certain algorithms and rules, the system will select a suitable replacement block from a set of predefined candidate blocks to replace the seed block.

[0033] Furthermore, the cycle block is formed by arranging and connecting the seed blocks according to a certain rule.

[0034] Furthermore, the edges of adjacent seed blocks are seamlessly connected to form a smooth transition effect during the tiling process.

[0035] Furthermore, the method for generating a four-sided continuous map from an arbitrary image includes the steps of: determining a loop region of appropriate size based on an input image; determining a seed block of the image according to the size of the loop region; extracting the best matching candidate block according to the seed block; continuing to fill according to the candidate block until the entire loop block is filled; and finally splicing all the loop blocks to ensure the best match at the splicing point to generate a final image.

[0036] Furthermore, the method for generating a four-dimensional continuous map from any picture comprises the steps of:

[0037] Determine the size of the loop region based on the input image, cut the original image into small blocks, and ensure that each small block can be tiled seamlessly; determine the seed block based on the loop region, and a good seed block should have good image continuity, that is, its edges can be seamlessly connected to form a visually smooth effect;

[0038] Determine candidate blocks according to the seed blocks, and set a transition area of ​​appropriate size at the edge when splicing the candidate blocks;

[0039] The transition area usually has obvious discontinuous parts. Compare the similarity of the two blocks, find the transition line with the largest difference, and use the Gaussian blur method of pixels to optimize the transition area.

[0040] Continue to fill in candidate blocks until the entire loop area is filled; after completing the above steps, if there are still discontinuous areas, you can set a frame selection area for secondary optimization.

[0041] The present invention also provides any of the above methods for generating a four-dimensional continuous map from any image for computer texture mapping.

[0042] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for generating a four-dimensional continuous map from any image, characterized in that: Includes cycle blocks, seed blocks and candidate blocks; The seed block is used as an independent image block; The cycle block is composed of a plurality of seed blocks and is used for seamless tiling in the map; The candidate blocks are candidate image blocks for replacing the seed blocks. In the process of generating the four-way continuous map, according to certain algorithms and rules, the system will select a suitable replacement block from a set of predefined candidate blocks to replace the seed block.

2. A method for generating a four-dimensional continuous map from any image according to claim 1, characterized in that: The cyclic blocks are formed by arranging and connecting the seed blocks according to a certain rule.

3. A method for generating a four-dimensional continuous map from any image according to claim 1, characterized in that: The edges of adjacent seed blocks are seamlessly connected to form a smooth transition effect during tiling.

4. A method for generating a four-dimensional continuous map from any image according to claim 1, characterized in that: The method comprises the following steps: determining a cycle area of ​​appropriate size based on an input image; determining a seed block of the image according to the size of the cycle area; extracting the best matching candidate block according to the seed block; continuing to fill the candidate block until the entire cycle block is filled; and finally splicing all the cycle blocks to ensure the best match at the splicing position and generate a final image.

5. A method for generating a four-dimensional continuous map from any image according to claim 4, characterized in that: Includes steps: Determine the size of the loop region based on the input image, cut the original image into small blocks, and ensure that each small block can be tiled seamlessly; determine the seed block based on the loop region, and a good seed block should have good image continuity, that is, its edges can be seamlessly connected to form a visually smooth effect; Determine candidate blocks according to the seed blocks, and set a transition area of ​​appropriate size at the edge when splicing the candidate blocks; The transition area usually has obvious discontinuous parts. Compare the similarity of the two blocks, find the transition line with the largest difference, and use the Gaussian blur method of pixels to optimize the transition area. Continue to fill in candidate blocks until the entire loop area is filled; after completing the above steps, if there are still discontinuous areas, you can set a frame selection area for secondary optimization.

6. A method for generating a four-dimensional continuous map from any image as claimed in any one of claims 1 to 5 for computer texture mapping.