An easy approach allowing automatic threshold computation on a restricted value range.

Motivation: A combination of the algorithms ‘automatic threshold’ and ‘multi-threshold segmentation’ is very powerful but occasionally yields unexpected results. Restricting the value range for the threshold computation can help.

Approach: Two segmentation steps are executed prior to the automatic threshold computation. These restrict the available value range for the threshold determination. Two customized algorithms are attached allowing either the generation of a new image object level – autosegmentation (creates new image object level).dcp – or the segmentation of a class on a preexisting image object level – autosegmentation (on existing image object level).dcp. In the short example both customized algorithms are combined to identify the body and the skeleton of a mouse from CT data.

• autosegmentation (creates new image object level).dcp(29 KB )
• Stack.zip(488 KB )
• Result.jpg(109 KB )
• autosegmentation (on existing image object level).dcp(29 KB )
• CT mouse analysis.dcp(65 KB )
• For downloading this file, please login or sign up for a new account.