The Local Otsu method is used in image processing to automatically perform clustering-based image thresholding. It reduces the gray scale image to a binary image. A global Otsu thresholding is available and explained here in the Image Segmentation section.

The Local Otsu method circumvents potential irregularities in the complete 3D image by focusing on smaller cubes with edge length in voxels according to the assigned Window Size.

The parameter Sample Length defines the distance between these 3D regions and can be set automatically by clicking Suggest Sample Length.

The Lower and Upper Cut Threshold can be adjusted to reduce noise in bright or dark regions. Depending on sample rate and window size, multiple thresholds are applied for assigning each voxel to the according phase.

Click Create Segmentation to segment the gray value image with the Local Otsu method. This thresholding method does not create a GeoDict result file. The resulting structure is displayed in the GeoDict visualization area. The gray value image stays in memory and can be compared to the structure. Check Structure and uncheck Volume Field in the view control panel above of the visualization area to only visualize the structure.

To save the structure in a *.gdt file select File → Save Structure as from the menu bar in the main GeoDict GUI.