Seamless stitching of tile scan microscope images.

Area of Science:

• Microscopy and Imaging
• Image Processing
• Computational Biology

Background:

• Optical imaging requires high-resolution, large-field-of-view images for biological diagnostics.
• Laser scanning microscopy often produces tile scans that need stitching into mosaics.
• Existing stitching methods fail with nonlinear, multimodal images, showing visible edge artefacts.

Purpose of the Study:

• To develop a novel method for seamlessly stitching image tiles from laser scanning microscopy.
• To eliminate visible edge artefacts and grey level jumps between tiles.
• To enable further image analysis on large-format microscopic images.

Main Methods:

• A novel stitching algorithm is presented to create seamless image mosaics.
• The algorithm corrects non-uniform brightness within individual tiles.
• It equilibrates neighbouring edges and common corner brightness to compensate for grey level differences.

Main Results:

• The developed technique successfully eliminates visible stitching artefacts.
• The corrected image mosaics are free from grey level jumps across tile boundaries.
• The method produces seamless mosaics suitable for subsequent image analysis.

Conclusions:

• The novel stitching method effectively addresses limitations of current techniques for laser scanning microscopy.
• The approach ensures seamless mosaic generation, crucial for accurate biological sample analysis.
• The technique is adaptable to other mosaic-assembly applications like astronomical or satellite imaging.