Two basic workflows are possible with Texture to Geometry:
To convert an assigned texture to a polygonal mesh
It is important that you select only a single surface when converting a texture to geometry. If the object is comprised of many grouped surfaces, you can use the Outliner or the Hypergraph to select a single surface that has the assigned texture.
The Texture to Geometry Options window appears.
Because it is possible to have multiple textures assigned to a surface, you must specify which image you want converted in the Input Image text box.
A polygonal mesh is created and placed above the texture on the selected polygon surface in the scene view. Whenever Fit to Selection is turned on (default), the resulting surface mesh is positioned above the selected surface in the scene view.
It’s not necessary to have a surface with an assigned texture in the scene in order to produce a polygonal mesh for an image. You can select the Texture to Geometry feature and specify an image in the options window, turn off the Fit to Selection option, and the resulting polygonal mesh will appear at the origin in the scene view.
To convert an image to a polygonal mesh
The Texture to Geometry Options window appears.
The resulting polygonal mesh is created and placed at the origin of the scene view along the X,Y plane.
The default settings for Texture to Geometry are a good starting point when converting an image/texture to a polygonal mesh. For specifics on what each option controls, see the Texture to Geometry options below.
Calculations for the segmentation phase of the feature are computational intensive and are affected by two factors.
The minimum image size should be no smaller than 256 x 256 pixels with an upper range of 4000 x 4000 pixels. Ideally, use input images in the range 400 to 512 pixels in one dimension. If your input image exceeds these guidelines you can resize the image using an image editing application prior to using the feature.
Texture to Geometry works well with images that contain well defined color features. For example, a bold colored logo with a clearly defined contrasting background is well suited for use with the feature.
Texture to Geometry does not work efficiently with images that have many random colors. For example, a noise texture would not be suitable. These types of images make it difficult for the segmentation algorithm to detect groups of related RGB color values that allow the feature to determine boundaries for subdivision of the polygonal mesh. In addition, images that contain subtle color artifacts as a result of image compression may be problematic. In these situations, it may be necessary to clean up the color artifacts in the image using an image editing application before using the Texture to Geometry feature.
For accurate fitting of the surface the UVs must be non-overlapping and should be within the 0 to 1 texture mapping range as they appear in the UV Texture Editor. When specifying an alternate UV set, the name of the UV set you specify must explicitly match the name of the UV set as it appears in the UV Texture Editor.
It is important that you select only a single surface for each conversion operation when converting a texture to geometry. If the object is comprised of many surfaces in a grouping, you can use the Outliner or the Hypergraph to select the single surface that has the assigned texture.