Focus measure through vectorization analysis for microscopic shape from focus

Area of Science:

• Microscopy
• Computational Imaging
• 3D Reconstruction

Background:

• Shape from Focus (SFF) reconstructs 3D shapes from multi-focus images.
• Traditional SFF struggles with micro-scale objects, complex surfaces, and limited fields of view.
• High magnification imaging presents challenges for accurate 3D surface reconstruction.

Purpose of the Study:

• To develop an advanced focus measure for precise 3D Shape from Focus (SFF).
• To overcome limitations of traditional methods in reconstructing micro-scale objects with complex surfaces.
• To enhance efficiency for large-scale 3D surface reconstruction.

Main Methods:

• Transformed grayscale images to gradient vector (GV) space.
• Applied divergence and curl operations in GV space to identify optimal focus regions.
• Developed strategies for temporal efficiency in large-scale scene reconstruction and multi-field image acquisition.

Main Results:

• The novel focus measure effectively handles complex surface structures and height mutations.
• Improved precision and robustness in 3D shape reconstruction of micro-scale objects.
• Enabled rapid surface shape computation for micro-scale objects within a short timeframe.

Conclusions:

• The proposed gradient vector-based focus measure significantly enhances Shape from Focus (SFF) capabilities.
• This method offers a robust solution for 3D reconstruction of challenging micro-scale surfaces.
• The strategies improve efficiency for extensive area reconstruction in micro-scale imaging.