A Model Hole Filling Method Based on Texture Synthesis

A texture synthesis and model technology, applied in 3D modeling, image data processing, instruments, etc., can solve the problem that the original geometric texture details of the model cannot be completely restored, the synthesis effect of geometric texture details is not ideal, and the three-dimensional texture cannot be simply satisfied Synthesis and other issues, to achieve the effect of enhancing operability and accuracy, optimizing the retention of detail features, and reducing the impact of errors

Active Publication Date: 2021-06-29
HUAQIAO UNIVERSITY
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Because most of the actual textures have rough surface geometry, image texture synthesis technology can no longer simply satisfy the synthesis of 3D textures; most researches focus on the theory and method improvement of image texture and geometric texture synthesis, such as choosing different Geometric texture synthesis method or adjust the sample matching criteria to improve the algorithm, but the synthesis effect of geometric texture details is not ideal
[0004] To sum up, since the existing mesh model hole repair methods still have problems such as low accuracy and the incomplete repair of the original geometric texture details of the model, it is necessary to propose a new solution for holes in textured mesh models.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Model Hole Filling Method Based on Texture Synthesis
  • A Model Hole Filling Method Based on Texture Synthesis
  • A Model Hole Filling Method Based on Texture Synthesis

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0135] The following demonstrates the above steps through specific examples:

[0136] Step 1, preprocessing process.

[0137] Specifically, input a textured hole model M (such as figure 2 shown), using the conventional method that does not consider the texture information to fill holes in the model to obtain the model M o (Such as image 3 shown), then to M o Perform smoothing operation to obtain the base surface mesh model M S (Such as Figure 4 shown), and then refer to the multi-resolution idea to separate the displacement vector representing the detailed feature information.

[0138] Step 2, base surface grid sampling.

[0139] Specifically, for the base mesh model M S Parametric processing is carried out, and the geodesic vector and coordinate base at each vertex are sequentially obtained by weighting and averaging the vertex-related information that has been parameterized by the grid DEM algorithm around the vertex, and then the improved discrete index mapping pa...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a model repairing method based on texture synthesis. A textured triangular mesh model containing holes is selected as input. First, holes are repaired in a conventional method without considering texture information on the input model, and smoothing is performed; secondly, the input model is smoothed by using Multi-resolution technology separates the characteristic displacement vector as the matching mark information; then parametrically processes the smoothed base surface mesh to obtain a two-dimensional plane, completes the sampling of each vertex on the parameterized plane, and obtains the area composed of each vertex sampling area The set of sample blocks; then search for the best matching sample block of the sample block to be repaired, and add the new texture information of the vertices to the base surface mesh, and update the vertex information in the area to be repaired sequentially similar to a spiral; finally according to the Update the grid vertex information, add constraints to solve the linear equations, reverse the new coordinates of the vertices to complete the target grid reconstruction. Therefore, the problem of missing texture features after repairing the textured triangular mesh model with holes is solved.

Description

technical field [0001] The invention belongs to the technical field of three-dimensional digital geometry processing, and in particular relates to a grid model repairing method. Background technique [0002] The hole repair of the mesh model is one of the three-dimensional digital geometry processing technologies. Factors such as three-dimensional scanning measurement equipment, measurement methods and human errors lead to the loss of model data. After point cloud preprocessing or triangulation, incomplete surface defects are formed. The formation of holes affects the visual effect of 3D models and post-geometry processing, such as surface modeling and editing, and the development of model repair technology also provides new means and tools for cultural relics preservation and research. The main research methods of hole repair include radial basis function method, diffusion method, B-spline curve interpolation method, template deformation method, etc., and hole repair itself...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G06T19/20G06T17/20G06T3/00
CPCG06T3/0037G06T17/20G06T19/20
Inventor 刘斌徐晓影许建文黄常标江开勇
Owner HUAQIAO UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products