NRStitcher

NRStitcher constructs deformation-aware image mosaics from large-scale microscopy data by detecting and correcting acquisition-induced deformations to produce artifact-free high-resolution assemblies.

Key Features:

• Deformation Detection and Correction: Employs local phase correlation to detect deformations between sub-images arising from environmental changes, radiation damage, or biological deterioration and applies corrections across the mosaic.
• Artifact-Free Mosaic Construction: Identifies and compensates for detected deformations to generate artifact-free, high-resolution image mosaics.
• Support for Distributed Computing: Implements support for distributed computing frameworks to process terabyte- to teravoxel-scale image mosaics across computing clusters.
• Benchmarking against Rigid Stitching: Has been benchmarked against existing rigid stitching implementations and reported to produce artifact-free mosaics with comparable runtime on the same hardware.

Scientific Applications:

• Neuroscience: Assembly of large-scale tomographic mosaics such as a 5.6 teravoxel reconstruction of mouse brain microvasculature.
• Developmental Biology and Organ Imaging: High-resolution stitching of complex structures, including microvasculature and whole-organ datasets, for structural and morphological analysis.

Methodology:

Detection and correction of deformations via local phase correlation applied to sub-images, and scalable processing by distributing computational tasks across nodes using distributed computing frameworks.