Improved texture view selecting and mapping method and system based on Markov random field

Abstract

The invention belongs to but not limited to the technical field of texture mapping, and discloses an improved texture view selecting and mapping method based on a Markov random field. The method comprises the following steps: firstly, determining texture images under different visual angles corresponding to each triangular surface in a grid model by adopting a graph cutting texture optimization algorithm based on a Markov random field; an optimal texture view is then selected for each triangular face. Triangular surface texture view selection is regarded as a multi-label problem, an energy function based on a Markov random field is constructed, data items are optimized for texture view selection, then a graph cut method is used for solving, and the optimal texture is selected for each triangular surface on the model. And finally, mapping the selected optimal texture to a triangular surface. According to the method, an image with clear texture and rich details can be selected for each triangular surface for texture mapping, and the authenticity of a reconstruction model is improved.

Description

Technical Field

[0001] The present invention belongs to but is not limited to the technical field of texture mapping, and in particular relates to an improved texture view selection and mapping method and system based on Markov random fields. Background Art

[0002] In the task of 3D reconstruction, the texture information of an object can be captured from multiple angles. This means that for each triangular face of the mesh model, the triangular face has a texture image at different viewing angles. When performing texture mapping, a triangular face often corresponds to multiple optional texture images. The texture image is affected by environmental factors such as shooting distance, shooting angle, ambient light, and anisotropy of object reflection, which will produce differences in color, brightness, etc., and may be out of focus, blurred, and color cast. Therefore, how to select the optimal texture from texture images at different viewing angles is the core challenge faced in high-quality 3D reconstruction. However, existing texture mapping techniques often fail to filter out the view with the richest texture information from a series of images, and are prone to produce obvious seams during the mapping process, which seriously reduces the authenticity of the reconstructed model.

[0003] In view of the above analysis, the technical problem that urgently needs to be solved in the prior art is: a texture view selection and mapping method is needed that can screen out views with the richest texture information from a series of images and produce no obvious seams during the mapping process. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides an improved texture view selection and mapping method and system based on Markov random field.

[0005] The present invention is implemented as follows: an improved texture view selection and mapping method based on Markov random field comprises the following steps:

[0006] Step 1, determining the texture images under different viewing angles corresponding to each triangular face in the mesh model;

[0007] Step 2: Select the optimal texture view for each triangular face. Treat the texture view selection of the triangular face as a multi-label problem, construct an energy function based on Markov random field, optimize the data item for texture view selection, and then use the graph cut method for optimization and solution.

[0008] Step three, map the selected optimal texture onto the triangular surface; use a combination of global texture optimization and local seam elimination to eliminate the color differences between texture images at different viewing angles.

[0009] Furthermore, the 3D mesh model consists of a triangular face set F = {F1, F2, …, F i} indicates that the texture image label used for texture mapping is I = {I1, I2, ..., I j}; The visibility of the triangular faces in the three-dimensional mesh model in the texture image is described by a matrix, the number of rows of the matrix is ​​equal to the number of triangular faces of the three-dimensional model, and the number of columns is equal to the number of texture images.

[0010] Furthermore, if a certain triangular face is visible in the texture image at a certain viewing angle, the element at the corresponding position of the matrix is ​​set to 1, and the texture image corresponding to the element 1 in each row is the texture image corresponding to the triangular face at different viewing angles.

[0011] Furthermore, the energy function has a data term and a smooth term, wherein the data term represents the possibility of selecting the optimal texture image for each triangular face, and the smooth term represents whether the texture image labels selected by adjacent triangular faces are consistent;

[0012] The energy function is as follows:

[0013]

[0014] The data items are as follows:

[0015]

[0016] In the formula, Grad ij The Sobel gradient operator is used to calculate the triangular surface F i Projection to image label l j The gradient on the image, the larger the gradient, the clearer the corresponding texture image, on the contrary, the blurrier the texture image; θ is the angle between the texture viewing angle and the normal of the triangle face, which is used to measure the consistency of the line of sight. If the angle between the line of sight and the normal vector is greater than 90°, the triangle face under this viewing angle is invisible and the texture image cannot be used; k is the color cast factor of the texture image, which is used to measure the degree of color cast of the texture image. When the k value is greater than 1.5, the texture image has color cast, otherwise, the texture image is normal.

[0017] The smoothing term is as follows:

[0018]

[0019] Among them, F i and F j Respectively represent two adjacent triangular faces, l i and l j Respectively represent F i and F j The texture image label to be selected. If l i and l j If they are equal, then F i and F jIf the two adjacent triangle faces select the same texture image, the smoothing term is 0, otherwise it is the maximum energy value.

[0020] Furthermore, the color cast factor k is calculated using an image color cast detection method based on an equivalent circle. The specific steps are as follows:

[0021] Step 1: Calculation of average chromaticity D of texture image

[0022]

[0023] Where M and N are the width and height of the texture image, M·N is the total number of pixels in the texture image, (d a ,d b ) is the center coordinate of the equivalent circle on the ab chromaticity plane, and D is the distance from the center of the equivalent circle to the origin of the neutral axis of the ab chromaticity plane (a=0, b=0).

[0024] Step 2: Calculation of chromaticity center distance M

[0025]

[0026] Where M and N are the width and height of the texture image, M·N is the total number of pixels in the texture image, and M is the radius of the equivalent circle on the ab chromaticity plane.

[0027] Step 3: Color cast assessment

[0028]

[0029] The color cast of the texture image is determined by the specific position of the equivalent circle on the ab chromaticity plane. When the obtained k value is not greater than 1.5, it is considered that the possibility of color cast of the texture image is small. a >0, reddish, otherwise greenish. b >0, yellowish, otherwise bluish.

[0030] Furthermore, in step three, a global texture optimization method based on color differences of mesh vertices is used to adjust the overall color of the texture, and a local texture optimization method based on Poisson editing is performed on the texture blocks to eliminate color differences and gaps between texture blocks.

[0031] Another object of the present invention is to provide an improved texture view selection and mapping method based on Markov random field and an improved texture view selection and mapping system based on Markov random field, comprising:

[0032] A texture image determination module determines the texture images under different viewing angles corresponding to each triangular face in the mesh model;

[0033] The optimal texture selection module selects the optimal texture view for each triangular face. The texture view selection of the triangular face is regarded as a multi-label problem, and an energy function based on Markov random fields is constructed to optimize the data items for texture view selection. Then, the graph cut method is used for optimization and solution.

[0034] The optimal texture mapping module maps the selected optimal texture to the triangular surface; it combines global texture optimization with local seam elimination to eliminate the color differences between texture images at different viewing angles.

[0035] Another object of the present invention is to provide a computer device, the computer device comprising a memory and a processor, the memory storing a computer program, and when the computer program is executed by the processor, the processor executes the steps of the improved texture view selection and mapping method based on Markov random field.

[0036] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to perform the steps of the improved texture view selection and mapping method based on Markov random fields.

[0037] Another object of the present invention is to provide an information data processing terminal, which includes the improved texture view selection and mapping system based on Markov random field.

[0038] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0039] First, the present invention adopts a graph-cut texture optimization algorithm based on Markov random fields. During texture mapping, one of the purposes of the energy function based on Markov random fields is to select the optimal texture image for mapping from multiple texture images at different viewing angles for each triangular face.

[0040] The present invention adopts a graph cut algorithm to optimize an energy function based on a Markov random field to obtain an optimal texture view corresponding to each triangular face.

[0041] The present invention uses a global texture optimization method based on color differences of mesh vertices to adjust the overall color of the texture, and then performs a local texture optimization method based on Poisson editing on the texture blocks to eliminate color differences and gaps between the texture blocks.

[0042] The present invention can select an image with clear texture and rich details for each triangular face for texture mapping, thereby improving the authenticity of the reconstructed model.

[0043] Second, the triangular faces of the mesh model have texture images at different perspectives. How to select high-quality textures from multiple texture images for mapping becomes a key challenge.

[0044] Solution: The present invention adopts a graph cut texture optimization algorithm based on Markov random fields. First, texture selection is regarded as a multi-label problem, and a global Markov model is constructed. When constructing the energy function, it is necessary to ensure the best texture quality. The data items of the present invention comprehensively consider information such as geometric perspective and image quality, introduce the gradient of the projection area of ​​the triangular face, the angle between the normal of the triangular face and the camera orientation, and the image color cast factor, and optimize the data items for texture selection. Then the graph cut method is used to solve and select the optimal texture view for each triangular face on the model.

[0045] Objects photographed from different perspectives have obvious color differences between the texture images due to changes in ambient lighting conditions, which causes color blocking and texture seams on the surface of the final 3D model, greatly reducing the color fidelity of the model.

[0046] Solution: A combination of global texture optimization based on mesh vertex color differences and local seam elimination based on Poisson editing is used to eliminate color differences between texture images at different viewing angles, ensure the tonal consistency and detail clarity of the model texture, and greatly enhance the realism of the 3D reconstructed model. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a flow chart of an improved texture view selection and mapping method based on Markov random field provided by an embodiment of the present invention;

[0048] Figure 2 is a structural diagram of an improved texture view selection and mapping system based on Markov random fields provided by an embodiment of the present invention;

[0049] Figure 3 is a schematic diagram of a texture-free 3D model provided by an embodiment of the present invention;

[0050] Figure 4 is a schematic diagram of a portion of texture images provided by an embodiment of the present invention;

[0051] Figure 5 is a schematic diagram of a texture mapping result without considering the color cast of an image provided by an embodiment of the present invention;

[0052] Figure 6 It is a schematic diagram of an improved texture mapping result provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0054] In multi-view texture mapping, each triangular face of the mesh model corresponds to a different texture image. It is necessary to select a clear texture and rich details image for each triangular face for texture mapping. In essence, this is an optimization problem under a Markov field, that is, the texture image labeling problem of the triangular face. Figure 1 As shown, an embodiment of the present invention provides an improved texture view selection and mapping method based on Markov random field, comprising the following steps:

[0055] Step 1, determining the texture images under different viewing angles corresponding to each triangular face in the mesh model;

[0056] Step 2: Select the optimal texture view for each triangular face. Treat the texture view selection of the triangular face as a multi-label problem, construct an energy function based on Markov random field, optimize the data item for texture view selection, and then use the graph cut method for optimization and solution.

[0057] Step three, map the selected optimal texture onto the triangular surface; use a combination of global texture optimization and local seam elimination to eliminate the color differences between texture images at different viewing angles.

[0058] The specific steps include:

[0059] Step 1: Determine the texture images under different viewing angles corresponding to each triangular face in the mesh model.

[0060] Assume that the 3D model consists of a triangular face set F = {F1, F2, ..., F i} indicates that the texture image label used for texture mapping is I = {I1, I2, ..., I j}. The visibility of the triangular faces in the 3D model in the texture image is described by a matrix. The number of rows in the matrix is ​​equal to the number of triangular faces in the 3D model, and the number of columns is equal to the number of texture images. If a triangular face is visible in the texture image at a certain viewing angle, the element at the corresponding position in the matrix is ​​set to 1. Since a triangular face is visible in multiple texture images, there are multiple positions in each row of the matrix that have elements set to 1. The texture image corresponding to the element set to 1 in each row is the texture image corresponding to the triangular face at different viewing angles.

[0061] During texture mapping, one of the purposes of the energy function based on Markov random fields is to select the optimal texture image for mapping from multiple texture images at different viewing angles for each triangular face.

[0062] Step 2: Select the optimal texture view for each triangular face. The texture view selection of the triangular face is regarded as a multi-label problem, and an energy function based on Markov random fields is constructed to optimize the data items for texture view selection, and then the graph cut method is used for optimization and solution.

[0063]

[0064] The energy function has a data term and a smoothness term. The data term represents the possibility of selecting the optimal texture image for each triangular face, and the smoothness term represents whether the texture image labels selected by adjacent triangular faces are consistent, which affects the texture image selection of adjacent faces.

[0065]

[0066] In the formula, Grad ij The Sobel gradient operator is used to calculate the triangular surface F i Projection to image label l j The gradient on the image, the larger the gradient, the clearer the corresponding texture image, on the contrary, the blurrier the texture image; θ is the angle between the texture viewing angle and the normal of the triangle surface, which is used to measure the consistency of the line of sight. If the angle between the line of sight direction and the normal vector is greater than 90°, the triangle surface under this viewing angle is invisible and the texture image cannot be used; k is the color cast factor of the texture image, which is used to measure the degree of color cast of the texture image. When the k value is greater than 1.5, the texture image has color cast, otherwise, the texture image appears normal. The present invention calculates the color cast factor k based on the image color cast detection method of the equivalent circle, and the specific steps are as follows:

[0067] Step 1: Calculation of average chromaticity D of texture image

[0068]

[0069] Where M and N are the width and height of the texture image, M·N is the total number of pixels in the texture image, (d a ,d b ) is the center coordinate of the equivalent circle on the ab chromaticity plane, and D is the distance from the center of the equivalent circle to the origin of the neutral axis of the ab chromaticity plane (a=0, b=0).

[0070] Step 2: Calculation of chromaticity center distance M

[0071]

[0072] Where M and N are the width and height of the texture image, M·N is the total number of pixels in the texture image, and M is the radius of the equivalent circle on the ab chromaticity plane.

[0073] Step 3: Color cast assessment

[0074]

[0075] The color cast of the texture image is determined by the specific position of the equivalent circle on the ab chromaticity plane. When the obtained k value is not greater than 1.5, it is considered that the possibility of color cast of the texture image is small. a >0, reddish, otherwise greenish. b >0, yellowish, otherwise bluish.

[0076] In addition, the smoothing term of the present invention is shown in formula (8).

[0077]

[0078] Among them, F i and F j Respectively represent two adjacent triangular faces, l i and l j Respectively represent F i and F j The texture image label to be selected. If l i and l j If they are equal, then F i and F j If the two adjacent triangle faces select the same texture image, the smoothing term is 0, otherwise it is the maximum energy value.

[0079] The graph cut algorithm is used to optimize the energy function based on Markov random field to obtain the optimal texture view corresponding to each triangle.

[0080] Step 3: Map the selected optimal texture view onto the triangular surface, and use a combination of global texture optimization and local seam elimination to eliminate the color differences between texture images at different viewing angles, making the final texture model more realistic.

[0081] After selecting and determining the optimal texture view corresponding to each triangular face, the selected optimal texture view is mapped to the triangular face.

[0082] When textures from different perspectives are spliced ​​in the same model, color blocking and texture seams will occur due to color differences. The present invention uses a global texture optimization method based on mesh vertex color differences to adjust the overall color of the texture, and then performs a local texture optimization method based on Poisson editing on the texture blocks to eliminate color differences and gaps between texture blocks.

[0083] like Figure 2As shown, an embodiment of the present invention provides an improved texture view selection and mapping method based on Markov random field and an improved texture view selection and mapping system based on Markov random field, comprising:

[0084] A texture image determination module determines the texture images under different viewing angles corresponding to each triangular face in the mesh model;

[0085] The optimal texture selection module selects the optimal texture view for each triangular face. The texture view selection of the triangular face is regarded as a multi-label problem, and an energy function based on Markov random fields is constructed to optimize the data items for texture view selection. Then, the graph cut method is used for optimization and solution.

[0086] The optimal texture mapping module maps the selected optimal texture to the triangular surface; it combines global texture optimization with local seam elimination to eliminate the color differences between texture images at different viewing angles.

[0087] 1. Specific application fields or related products of the present invention

[0088] The present invention is mainly used in the field of digital cultural relic protection and restoration, and digitally models and archives cultural relics through 3D reconstruction and texture mapping technology. This technology can not only be used for display and storage, but also provide reference for actual restoration, such as ancient sculptures, murals and other cultural relics with complex textures.

[0089] An important application of cultural relics digitization is to conduct comprehensive digital archiving of historical relics, providing a basis for future restoration and reproduction of cultural relics. In the process of digital archiving, the present invention can significantly improve the authenticity and integrity of the digital model through an optimized texture mapping algorithm, making the digital archiving more accurate.

[0090] For artifacts with complex surface textures, such as ancient ceramics, metal utensils, and sculptures, the optimized texture mapping technology of the present invention can restore details and improve visual authenticity. This provides precise technical support for digital restoration and avoids the problems of detail loss and color distortion in traditional methods.

[0091] The high-precision three-dimensional digital model generated by the present invention can be applied to virtual display, education and communication in museums. By optimizing the texture, the color and details of the model are closer to the real cultural relics, providing a more intuitive and realistic experience for the audience.

[0092] II. Evidence of the Technical Effects Obtained by the Embodiments of the Present Invention

[0093] The present invention solves the image color cast problem in traditional methods by introducing an image color cast factor in texture mapping. Experiments show that the color cast problem can cause the model to have obvious differences in color from the actual object, such as the distortion of the eye color of the model dog in the experiment. The optimization method of the present invention successfully eliminates the color cast, making the model color closer to the real object.

[0094] Traditional texture mapping methods are prone to detail loss in areas with rich details, such as the skin folds on the belly of the model dog, resulting in a lack of realism in the final model. This invention comprehensively considers the detailed information of the image when optimizing texture selection, significantly improving the visual realism in these detailed areas and making the model texture more delicate.

[0095] The optimized energy function of the present invention not only considers color consistency, but also makes improvements specifically for detail recovery. In the process of selecting the optimal texture of the triangle surface, by combining multi-angle images, it is possible to effectively select textures with complete details and accurate colors. Experimental results show that this method is superior to traditional methods in visual effects.

[0096] Comparative experiments show that traditional texture mapping methods are prone to color discontinuity or inconsistency in complex texture areas. The present invention solves the color consistency problem through an optimization algorithm, making the generated model more visually attractive, especially in detail areas such as the abdomen and head of the model dog, where the optimized color and texture transitions are more natural.

[0097] Multiple Views on Complex Artifacts Figure 3 In the 3D reconstruction, the algorithm of the present invention significantly improves the quality and efficiency of texture mapping. By extracting the optimal texture from multi-angle images and eliminating the problem of splicing seams, the generated digital model of cultural relics is complete and realistic, providing a reliable basis for actual restoration and analysis.

[0098] Experimental results show that the present invention has obvious advantages over traditional methods in color correction and detail restoration. Both the fine texture of the surface of cultural relics and the overall visual quality are closer to real objects, providing an efficient and accurate solution for digital applications in the field of cultural relics protection. This technological breakthrough not only enhances the visual appeal of digital models, but also provides an important reference for subsequent cultural relics restoration.

[0099] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. It can be understood by a person of ordinary skill in the art that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0100] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. An improved texture view selection and mapping method based on Markov random field, characterized in that: The following steps are involved: Step 1, determining the texture images under different viewing angles corresponding to each triangular face in the mesh model; Step 2: Select the optimal texture view for each triangle face; The texture view selection of the triangle surface is regarded as a multi-label problem, and an energy function based on Markov random field is constructed to optimize the data item for texture view selection, and then the graph cut method is used for optimization and solution. Step 3, mapping the selected optimal texture onto the triangle surface; A combination of global texture optimization and local seam elimination is used to eliminate color differences between texture images at different viewing angles.

2. The improved Markov random field-based texture view selection and mapping method as claimed in claim 1, characterized in that: The 3D mesh model consists of a triangular face set F = {F1, F2, …, F i } indicates that the texture image label used for texture mapping is I = {I1, I2, ..., I j }; The visibility of the triangular faces in the three-dimensional mesh model in the texture image is described by a matrix, the number of rows of the matrix is ​​equal to the number of triangular faces of the three-dimensional model, and the number of columns is equal to the number of texture images.

3. The improved Markov random field-based texture view selection and mapping method as claimed in claim 2, characterized in that: If a certain triangular face is visible in the texture image at a certain viewing angle, the element at the corresponding position of the matrix is ​​set to 1. The texture image corresponding to the element 1 in each row is the texture image corresponding to the triangular face at different viewing angles.

4. The improved Markov random field-based texture view selection and mapping method as claimed in claim 1, characterized in that: The energy function has a data term and a smooth term, where the data term represents the possibility of selecting the optimal texture image for each triangular face, and the smooth term represents whether the texture image labels selected by adjacent triangular faces are consistent; The energy function is as follows: The data items are as follows: In the formula, Grad ij The Sobel gradient operator is used to calculate the triangular surface F i Projection to image label l j The gradient on the image, the larger the gradient, the clearer the corresponding texture image, on the contrary, the blurrier the texture image; θ is the angle between the texture viewing angle and the normal of the triangle surface, which is used to measure the consistency of the line of sight. If the angle between the line of sight direction and the normal vector is greater than 90°, the triangle surface under this viewing angle is invisible and the texture image cannot be used; k is the color cast factor of the texture image, which is used to measure the degree of color cast of the texture image; when the k value is greater than 1.5, the texture image has color cast, otherwise, the texture image is normal; The smoothing term is as follows: Among them, F i and F j Respectively represent two adjacent triangular faces, l i and l j Respectively represent F i and F j The selected texture image label; if l i and l j If they are equal, then F i and F j If the two adjacent triangle faces select the same texture image, the smoothing term is 0, otherwise it is the maximum energy value.

5. The improved Markov random field-based texture view selection and mapping method as claimed in claim 4, characterized in that: The color cast factor k is calculated using the image color cast detection method based on the equivalent circle. The specific steps are as follows: Step 1: Calculation of average chromaticity D of texture image Where M and N are the width and height of the texture image, M·N is the total number of pixels in the texture image, (d a ,d b ) is the center coordinate of the equivalent circle on the ab chromaticity plane, and D is the distance from the center of the equivalent circle to the origin of the neutral axis of the ab chromaticity plane (a=0, b=0); Step 2: Calculation of chromaticity center distance M Where M, N are the width and height of the texture image, M·N is the total number of pixels of the texture image, and M is the radius of the equivalent circle on the ab chromaticity plane; Step 3: Color cast assessment The color cast of the texture image is determined by the specific position of the equivalent circle on the ab chromaticity plane; when the obtained k value is not greater than 1.5, it is considered that the possibility of color cast of the texture image is small; d a >0, reddish, otherwise greenish; d b >0, yellowish, otherwise bluish.

6. The improved Markov random field based texture view selection and mapping method as claimed in claim 1, characterized in that: In the step three, a global texture optimization method based on color differences of mesh vertices is used to adjust the overall color of the texture, and a local texture optimization method based on Poisson editing is performed on the texture blocks to eliminate color differences and gaps between the texture blocks.

7. An improved Markov random field based texture view selection and mapping system according to any one of claims 1 to 6, comprising: A texture image determination module determines the texture images under different viewing angles corresponding to each triangular face in the mesh model; The optimal texture selection module selects the optimal texture view for each triangle face; The texture view selection of the triangle surface is regarded as a multi-label problem, and an energy function based on Markov random field is constructed to optimize the data item for texture view selection, and then the graph cut method is used for optimization and solution. The optimal texture mapping module maps the selected optimal texture to the triangular surface; it combines global texture optimization with local seam elimination to eliminate the color differences between texture images at different viewing angles.

8. A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the improved texture view selection and mapping method based on Markov random field as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to perform the steps of the improved texture view selection and mapping method based on Markov random fields as claimed in any one of claims 1 to 6.

10. An information data processing terminal, comprising the improved Markov random field-based texture view selection and mapping system as claimed in claim 7.

Images