Microscope Camera Pipeline (Arducam + Raspberry Pi)

Lightweight camera pipeline for capturing microscope slide images using an Arducam connected to a Raspberry Pi, then automatically cropping the region of interest for downstream analysis.

Overview

This project captures an image from an Arducam camera mounted on a microscope and applies basic computer vision to detect the slide area and crop the image automatically. The goal is to reduce manual preprocessing and produce consistent, analysis-ready images.

Features

• Image capture from Arducam on Raspberry Pi
• Automatic slide/ROI detection via contour-based processing
• Cropped output images for analysis or upload

Project Structure

• ContourTest1.py
  Python script for contour detection and auto-cropping
• takePhoto.sh
  Shell script to trigger image capture from the camera
• install_pivariety_pkgs.sh
  Installs Arducam / Pi camera dependencies

Setup

Tested on Raspberry Pi with an Arducam camera.

# Install camera dependencies
bash install_pivariety_pkgs.sh

# (Optional) create virtual environment
python3 -m venv venv
source venv/bin/activate

# Install Python dependencies
pip install opencv-python numpy

Usage

# Capture an image from the camera
bash takePhoto.sh

# Run auto-cropping pipeline
python3 ContourTest1.py

Processed images are saved locally.

Notes

• Lighting conditions significantly affect contour detection accuracy

• Camera alignment and focus matter for reliable cropping

• This is a task-specific pipeline, not a general-purpose vision framework

Future Work

• More robust slide edge detection under uneven lighting

• Batch processing for multiple images

• Optional cloud upload integration

Disclaimer

This repository contains experimental scripts and is intended for prototyping and coursework.