An Image-Based 3D Quadrilateral Mesh Model Reconstruction Method

Abstract

The invention discloses an image-based three-dimensional quadrilateral grid model reconstruction method, comprising: step (1): acquiring point cloud data representing the shape of an image object or scene, and recovering the three-dimensional information of the image; step (2): constructing a point cloud The four-sided grid of the data, to reconstruct the four-sided grid model. It has the advantage of fitting better image objects or scenes.

Description

technical field

[0001] The invention relates to an image-based three-dimensional quadrilateral mesh model reconstruction method. Background technique

[0002] With the continuous development of computer software and hardware technology, real-time rendering of large-scale complex scenes has become possible, which puts forward new requirements for the complexity and realism of the model. Although traditional 3D modeling tools are improving day by day, building complex 3D models is still a very time-consuming and laborious task. Since many 3D models can be found or modeled in the real world, image-based modeling techniques have become the ideal modeling method in people's minds. In addition, traditional 3D modeling generally can only obtain the geometric information of the scene, while image-based 3D modeling provides a natural way to generate photo-realistic synthetic images, and has become a hot spot in the field of computer graphics. , It is widely used in digital preserva...

Examples

specific Embodiment

[0087] Such as figure 1 Shown, the specific embodiment that the present invention provides is as follows:

[0088] An image-based three-dimensional quadrilateral mesh model reconstruction method, comprising:

[0089] Step (1): Obtain point cloud data representing the shape of the image object or scene, and restore the three-dimensional information of the image;

[0090] Step (2): Construct the quadrilateral grid of the point cloud data, and reconstruct the quadrilateral grid model.

[0091] The step of described step (1) is:

[0092] Step (11): Obtain the three-dimensional information of the matching point through the self-calibration of the camera and the feature matching of the image;

[0093] Step (12): Construct a neural network, train the neural network with matching points and their three-dimensional coordinates as samples, and obtain point cloud data of image objects or scene shapes.

[0094] The step of described step (11) is:

[0095] Step (111): Let the original i...

technical field

[0001] The invention relates to an image-based three-dimensional quadrilateral mesh model reconstruction method. Background technique

[0002] With the continuous development of computer software and hardware technology, real-time rendering of large-scale complex scenes has become possible, which puts forward new requirements for the complexity and realism of the model. Although traditional 3D modeling tools are improving day by day, building complex 3D models is still a very time-consuming and laborious task. Since many 3D models can be found or modeled in the real world, image-based modeling techniques have become the ideal modeling method in people's minds. In addition, traditional 3D modeling generally can only obtain the geometric information of the scene, while image-based 3D modeling provides a natural way to generate photo-realistic synthetic images, and has become a hot spot in the field of computer graphics. , It is widely used in digital preserva...

Examples

specific Embodiment

[0087] Such as figure 1 Shown, the specific embodiment that the present invention provides is as follows:

[0088] An image-based three-dimensional quadrilateral mesh model reconstruction method, comprising:

[0089] Step (1): Obtain point cloud data representing the shape of the image object or scene, and restore the three-dimensional information of the image;

[0090] Step (2): Construct the quadrilateral grid of the point cloud data, and reconstruct the quadrilateral grid model.

[0091] The step of described step (1) is:

[0092] Step (11): Obtain the three-dimensional information of the matching point through the self-calibration of the camera and the feature matching of the image;

[0093] Step (12): Construct a neural network, train the neural network with matching points and their three-dimensional coordinates as samples, and obtain point cloud data of image objects or scene shapes.

[0094] The step of described step (11) is:

[0095] Step (111): Let the original i...

Images