Volume Segmenter

(Not recommended) Segment 3-D grayscale or RGB volumetric images

Volume Segmenter is not recommended (since R2026a). Use the Medical Image Labeler (Medical Imaging Toolbox) app instead. For more information, see Compatibility Considerations.

Description

The app can be used to create and refine a binary or semantic segmentation mask for a 3-D grayscale or an RGB image using automated, semi-automated, and manual techniques.

Note

The app is not supported in MATLAB^® Online™. For details, see Specifications and Limitations.

Open the Volume Segmenter App

• MATLAB Toolstrip: Open the Apps tab, under Image Processing and Computer Vision, click the Volume Segmenter app icon.

• MATLAB command prompt: Enter volumeSegmenter.

Examples

Load Volume and Labeled Volume into Volume Segmenter

Load a volume into the workspace.

load(fullfile(toolboxdir("images"),"imdata","BrainMRILabeled", ...
    "images","vol_001.mat"));

Load the corresponding labeled volume into the workspace.

load(fullfile(toolboxdir("images"),"imdata","BrainMRILabeled", ...
    "labels","label_001.mat"));

Open Volume Segmenter specifying both the volume and the labeled volume.

volumeSegmenter(vol,label)

Related Examples

• Create Binary Mask Using Volume Segmenter
• Create Semantic Segmentation Using Volume Segmenter
• Work with Blocked Images Using Volume Segmenter

Programmatic Use

volumeSegmenter opens the Volume Segmenter app.

volumeSegmenter(V) opens the Volume Segmenter app, loading the volume V into the app. Volume V is an m-by-n-by-p or m-by-n-by-p-by-3 image of data type uint8, uint16, uint32, int8, int16, int32, single, or double.

volumeSegmenter(V,L) opens the Volume Segmenter app, loading the volume V and the labeled volume L into the app. The labeled volume L is an m-by-n-by-p image of data type logical, categorical, uint8, uint16, uint32, int8, int16, int32, single, or double.

volumeSegmenter(___,Show3DDisplay=TF) logical value that specifies whether Volume Segmenter includes a visualization of the 3-D volume in the app. The default value is true. Hiding the 3-D display by specifying Show3DDisplay as false can improve app performance on platforms with limited graphics support, such as Linux platforms or Windows platforms that use software rendering.

Version History

Introduced in R2020b

R2026a: Not recommended

The Volume Segmenter app is no longer recommended. Use the Medical Image Labeler (Medical Imaging Toolbox) app instead (requires Medical Imaging Toolbox™). Medical Image Labeler offers several advantages, such as:

• Use of spatial referencing information found in file metadata.

• Simultaneous display of three orthogonal cross-sections.

• Improved performance.

R2021a: Support for blocked images

The Volume Segmenter app now accepts blockedImage objects as input for segmentation. Use blocked images when a volume is too large to fit into memory. By using blocked images, you can segment volumes without running out of memory.

After loading the blocked image into Volume Segmenter, you can use the same set of tools the app offers for nonblocked images to segment regions of the volume. This includes the use of automated methods, working block-by-block through each slice. Use the Blocked Image tab in Volume Segmenter to specify the block you want to work on, and to mark that block completed. Volume Segmenter automatically calculates the percent of blocks that are done. Volume Segmenter combines your work across blocks to create masks. For more information, see Work with Blocked Images Using Volume Segmenter.