This article describes the basic Workflow for building 3D model of a object. We also recommend to think through the process of photography. You can found more detailed recommendations in the section - Image Capture Tips.
1. To add photos select Add Photos... command from the Workflow menu.
2. In the Add Photos dialog browse the source folder and select files to be processed.
Note: Recommended to mask all irrelevant elements on the source photos (background, accidental, foreground, etc.). You can find more information in tutorial - Masks.
At this stage Metashape estimates the camera position for each photo and builds the sparse point cloud.
1. Select Align Photos command from the Workflow menu.
2. Set following recommended values for the parameters in the Align Photos dialog:
- If you have a large number of images (more than 100), we recommend to enable - Generic preselection option, to reduce processing time;
- We do not recommend using parameters: Guided image matching, Adaptive camera model fitting;
- For Key point limit the recommended range of 20 000 to 100 000;
- For Tie point limit the recommended range of 2 000 to 40 000.
3. Click OK button to start photo alignment.
After photo alignment is finished, refine bounding box position and orientation to fit the desired volume for reconstruction:
This step is optional since Metashape calculates bounding box dimensions, orientation and location automatically. But it is recommended to check if any correction is needed, because further reconstruction step deals only with the data inside the volume. Bounding box can be resized, moved and rotated with the help Resize Region, Move Region and Rotate Region tools.
Build Dense Cloud
Based on the estimated camera positions the program calculates depth information for each camera to be combined into a single dense point cloud.
1. Select Build Dense Cloud command from the Workflow menu. Set the following recommended values for the parameters in the Build Dense Cloud dialog:
- We recommend to use - Medium quality. Higher quality takes quite a long time and demands more computational resources, but allows for more detailed results generation.
- If the geometry of the scene to be reconstructed is complex with numerous small detail on the foreground, then it is recommended to set Mild depth filtering mode, for important features not to be sorted out).
2. Click OK button.
Note: Points from the dense cloud can be removed with the help of selection tools and Delete/Crop instruments located on the Toolbar.
After dense point cloud has been reconstructed it is possible to generate polygonal mesh model based on the dense cloud data or depth maps data. The second approach supports GPU-acceleration and mostly provides better results for the same depth maps quality for objects and scenes with big number of minor details.
1. Select Build Mesh command from the Workflow menu. Set the following recommended values for the parameters in the Build Mesh dialog:
- If at the previous stage - Depth maps was built and save in project, we recommend using the parameter - Reuse depth maps.
2. Click OK button to start building geometry.
1. Select Build Texture command from the Workflow menu.
Set the following recommended values for the parameters in the Build Texture dialog:
2. Click OK button to start building texture.
- Usually the texture atlas size in range of 4000 - 10000 pixels is sufficient.
1. In case the model should be exported to the file, select Export Model... command from the File menu.
2. In the Export Model... in the dialog box, select options to export the 3D model.
3. Click OK button.
Note: In the Export Model dialog indicate the desired export parameters. Note that the list of available exporting parameters depends on the selected file format. In the example *.fbx format was selected.
To publish your model online use Upload Data... command from the File menu.