generate

Generate 3-D asset from single view camera image

Since R2025a

Syntax

[assetAttributes,info] = generate(assetGenerator,I)

[___] = generate(___,Name=Value)

Description

Add-On Required: This feature requires the Scenario Builder for Automated Driving Toolbox add-on.

[assetAttributes,info] = generate(assetGenerator,I) generates a 3-D asset from a single view camera image, I, using an imageAssetGenerator object, assetGenerator. The function returns the 3-D asset attributes, assetAttributes. Additionally, the function returns the debugging information, info, which also contains the preprocessed image.

[___] = generate(___,Name=Value) specifies options using one or more name-value arguments in addition to the arguments from the previous syntax. For example, ExecutionEnvironment="cpu" uses the CPU as the hardware resource to execute the function.

Note: This feature also requires the Deep Learning Toolbox™.

Examples

Generate 3D Asset from Single View Camera RGB Image

Read a single view camera RGB image into the workspace.

I = imread("ObjaverseAssetImage.jpeg");
imshow(I)

Initialize the image asset generator object.

assetGenerator = imageAssetGenerator;

Generate an asset from the image by using the generate object function of the imageAssetGenerator object.

assetAttributes = generate(assetGenerator,I);

Create a surface mesh from the obtained asset attributes.

assetMesh = surfaceMesh(assetAttributes.vertices,assetAttributes.faces,VertexColors=assetAttributes.vertexColors);

Display the created surface mesh.

surfaceMeshShow(assetMesh,BackgroundColor="White")

Input Arguments

assetGenerator — 3-D asset generator

imageAssetGenerator object

3-D asset generator, specified as an imageAssetGenerator object.

I — Input RGB image

M-by-N-by-3 numeric array

Input RGB image, specified as an M-by-N-by-3 numeric array.

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Example: generate(assetGenerator,I,ExecutionEnvironment="cpu") uses the CPU as the hardware resource to generate 3-D assets.

MeshResolution — Resolution of 3-D asset mesh

256 (default) | 32 | 64 | 128 | 512 | 1024

Resolution of the 3-D asset mesh, specified as 32, 64, 128, 256, 512, or 1024. The MeshResolution value represents the number of grids along each axis of the 3-D asset. Increase this value to generate a more detailed mesh. Increasing the value results in increased memory consumption and higher computation time.

ExecutionEnvironment — Hardware resource

"auto" (default) | "gpu" | "cpu"

Hardware resource, specified as "auto", "gpu", or "cpu".

• "auto" — Use a GPU if one is available. Otherwise, use the CPU. The use of a GPU requires a Parallel Computing Toolbox™ license and a CUDA^® enabled NVIDIA^® GPU. For information about the supported compute capabilities, see GPU Computing Requirements (Parallel Computing Toolbox).

• "gpu" — Use the GPU. If a suitable GPU is not available, the function returns an error message.

• "cpu" — Use the CPU.

Output Arguments

assetAttributes — 3-D asset attributes

structure

3-D asset attributes, returned as a structure. The structure contains these fields:

• Vertices — Mesh vertices, returned as a P-by-3 matrix. Each row of the matrix is of the form [x y z], specifying the coordinates of a vertex. Each vertex has a vertex ID equal to its row number in the matrix. P is the total number of vertices in the mesh.

• Faces — Mesh triangular faces, returned as a Q-by-3 matrix. Each row of the matrix is of the form [V₁ V₂ V₃], specifying the row indices of the corresponding vertices in Vertices that define the triangular face. Q is the number of faces in the mesh.

• VertexColors — Color values for the mesh vertices, returned as a P-by-3 matrix. Each row of the matrix is of the form [R G B], specifying the color value for a vertex. Each value must be in the range [0, 1]. P is the total number of vertices in the mesh.

info — Debug information

structure

Debug information, returned as a structure. The structure has a preProcessedImage field that contains the preprocessed input image. The size of the preprocessed image and the size of the input image are the same.

Version History

Introduced in R2025a