Deinterlacing

This python module can be used to correct the misalignment between forward and backward-scanned lines collected by resonance-scanning microscopes.

Features

• GPU Acceleration: Optional CuPy backend for increased performance
• Batch Processing: Supports block-wise processing to reduce memory constraints.
• Pooling Noisy Data: Deinterlacing can be applied to pooled-pixels for improved performance on noisy or sparse images.
• Handles Instability: Supports processing individual frames while autocorrection methods applied during acquisition stabilize
• Sub-Pixel: Pixel & Sub-Pixel registration available

Installation

The repository is available on PyPI and can be installed using your preferred package manager. For example: pip

pip install deinterlacing

uv

uv add deinterlacing

Dependencies

• Boltons
• CuPy (Optional)
• NumPy
• Pydantic
• TQDM

Example

from deinterlacing import deinterlace
import numpy as np

# Load your images
images = np.load("my_images.npy")

# Deinterlace the images
deinterlace(images)