1. Image alignment for 3D reconstruction in a SEM
- Author
-
Simona Podda, Ruggero Pintus, and Massimo Vanzi
- Subjects
Channel (digital image) ,Computer science ,business.industry ,Computation ,Instrumentation ,3D reconstruction ,Detector ,Condensed Matter Physics ,Translation (geometry) ,Atomic and Molecular Physics, and Optics ,Surfaces, Coatings and Films ,Electronic, Optical and Magnetic Materials ,Depth map ,Computer vision ,Artificial intelligence ,Electrical and Electronic Engineering ,Safety, Risk, Reliability and Quality ,business ,Rotation (mathematics) - Abstract
In a previous paper [1] a procedure to extract the third dimension from back-scattered electron has been proposed. The 3D recovery process involves the acquisition of four images and the computation of surface gradients; thus, a specific algorithm finds the depth map of the given surface. The image acquisition process may require or specific instrumentation (4-sector, independent channel axial BS detector), or very standard SEM configuration and specimen rotation/translation. In this last case an alignment procedure should be developed, which results troublesome because of the different image shadowing. The paper deals with an alignment method based on the parallel acquisition of the BS images and of their respective uniformly shaded twins obtained by the specimen current detector, or by the standard annular axial BS detector.
- Published
- 2005
- Full Text
- View/download PDF