The Watershed Segmentation algorithm is a region based segmentation algorithm that segments a single image layer of a scene. Local intensity minima are used as seed objects, and the objects grow with rising intensity levels into the neighborhood until they touch objects growing from neighboring seeds. An intuitive visualization is that of water rising from intensity valleys until the whole area is “flooded”. Object borders are formed in places where the rising water levels from different valleys meet.
Learn more here: Watershed segmentation video
This algorithm is an alternative to the Multiresolution Segmentation algorithm and supports at the moment only small images but no videos.
Supported Domains
Pixel Level; Image Object Level; Current Image Object; Neighbor Image Object; Super Object; Sub Objects; Linked Objects
Algorithm Parameters
Level
Available only if the domain is pixel level. Select an existing level or enter a name.
Layer
Select the input image layer.
Invert layer
Specify whether to invert the image layer. Default setting is No. If set to No dark pixels will be used as seeds, if set to Yes bright pixels will be used as seeds.
Neighborhood
Choose between 4 - connected, 8 - connected or 8-disconnected pixel connectivity.
Default setting is 8 - connected.
Specify whether seeds can grow into the four neighboring pixels (left, right, top, bottom) or into 8 neighboring pixels (i.e. they can grow diagonally). In the latter case, you can decide whether final objects will be converted to connected objects (8- connected) or not (8-disconnected). See also Pixel Neighborhood and Connectivity of Image Objects.
Seed based fusion
Seed criterion
With the default behavior, each initial seed (i.e. each local intensity minimum) will correspond to an object of the final segmentation. However, it is possible to define certain criteria under which two initially distinct seeds will be merged. The criterion is always evaluated when two objects first “touch” each other, i.e., they start having a common border.
- Ovfl. Area in pixels: seeds whose area is below the criterion will be merged into neighboring objects/seeds
- Ovfl. Height: seeds whose maximum intensity is below the criterion will be merged into neighboring objects/seeds
- Ovfl. Depth: seeds whose intensity range (difference between maximum and minimum intensity) are below the criterion will be merged into neighboring obejcts/seeds
- Ovfl. Volume: for each pixel, the difference to the current intensity maximum is evaluated and the sum of these values corresponds to the “volume” (think: amount of “water” in the “valley”). If this volume is below the criterion, seeds will be merged into neighboring objects/seeds
Up to three seed criteria can be selected and combined.
Threshold
Select a threshold for the seed criterion.
Fuse super ojects
Specify whether to fuse super object. Default setting is No.
