How to Segment Cavities Using the Compute Ambient Occlusion Module
This tutorial demonstrates how to segment pores or cavities using the Compute Ambient Occlusion module.
The Compute Ambient Occlusion module enables the segmentation of pores and cavities that usually cannot be distinguished from the surrounding material by considering the grey values and texture properties of the image.
This video first explains how the Compute Ambient Occlusion module comes in handy to segment cavities that are connected to, and have the same intensity level as, the background outside the object, making it impossible to use the traditional thresholding methods.
Next, the video briefly explains how the ambient occlusion is computed and can then be used to identify the regions that are inside the object.
Finally, the method is demonstrated on a mouse spine dataset, where the pore space is segmented.
View the trabecular and cortical bone segmentation tutorial to see how the Compute Ambient Occlusion can be used to further segment the trabecular part of the mouse spine, and finally aggregate in one label field the pore space, trabecular part, and cortical part of the spine.
Additional resources:
Courtesy of National Laboratory Animal Center – National Applied Research Laboratories, Taiwan.