Human facial attribution via Raspberry Pi-based Video Surveillance with Low-Cost Webcams

Features extracted

• Gender
• Age
• Emotion

Check the RPi_Mini_Project.pdf for more details

Check the Main prepended files(s) for final execution in each of the Age_Gender_Detection or Emotion_Detection directories

Backup-Restore environment Script Usage Guide

Check the backups directory

1. Restore on Another System:

bash restore_env.sh

2. Backup Current Environment:

bash backup_env.sh

---

🧠 Things to Watch Out For while restoring the environment:

Issue	How to Handle
Different Raspberry Pi OS versions	Prefer using the same OS version (e.g., Raspberry Pi OS Lite 64-bit).
Packages needing compilation (e.g., OpenCV)	Either install from apt or recompile carefully.
Python version mismatch	Install the matching version via pyenv or source compilation.

🛠️ In Short, Your Reproduction Workflow:

• Install Python and pip.
• Create a virtual environment with --system-site-packages.
• Activate the virtual environment.
• Install packages from venv-requirements.txt using pip install -r venv-requirements.txt.
• Reinstall system packages if required (e.g., OpenCV).
• Restore environment variables if applicable.
• Verify the setup by checking Python version and installed packages.

Prefer manual backup and restoration? Expand for detailed steps!

Backup steps

✨ Step 1: List venv installed packages
Normal:

pip freeze > venv-requirements.txt

✅ Captures whatever was installed directly into venv.

---

✨ Step 2: List system-site-packages (that your venv can see)
First, find where the system-site-packages are located:

python -c "import site; print(site.getsitepackages())"

Example output:

['/usr/lib/python3.10/site-packages', '/usr/local/lib/python3.10/dist-packages']

Then, list all packages installed there using:

pip list --path /usr/lib/python3.10/site-packages > system-site-packages.txt
pip list --path /usr/local/lib/python3.10/dist-packages >> system-site-packages.txt

✅ This way you get everything available via system-site-packages separately.

---

✨ Step 3: Capture environment variables too (optional)
This sometimes matters because OpenCV can behave differently based on LD_LIBRARY_PATH, PATH, etc.

printenv > environment-variables.txt

✅ Useful for reproducing runtime behavior.

---

✨ Step 4: Save Python Version
Remember: small version differences (3.10.12 vs 3.10.14) can break things like OpenCV.

Save:

python --version > python-version.txt

✅ Ensures you know the exact Python version needed.

---

✨ Step 5: Final backup
You now have:

venv-requirements.txt
system-site-packages.txt
environment-variables.txt
python-version.txt

Put all of these into a folder and zip it:

mkdir env-backup
mv venv-requirements.txt system-site-packages.txt environment-variables.txt python-version.txt env-backup/
tar -czvf env-backup.tar.gz env-backup

---

Restore steps

[Assuming you have the env-backup.tar.gz saved.]

✅ Step 1: Setup the Target System
First, on your new Raspberry Pi or any Linux machine:

Install Python (same version as saved in python-version.txt):

sudo apt update
sudo apt install python3 python3-pip python3-venv

If you want the exact version (say Python 3.9.2), you may need to install manually or via pyenv.

Install pip if not already installed:

sudo apt install python3-pip

---

✅ Step 2: Create a Virtual Environment (with system-site-packages)
Remember you originally created the venv with --system-site-packages?
So recreate it exactly:

python3 -m venv --system-site-packages myenv

✅ This ensures your venv can also see the system's pre-installed packages (like OpenCV, etc.).

---

✅ Step 3: Activate the Virtual Environment

source myenv/bin/activate

---

✅ Step 4: Restore Python Packages
Assume you have your backup archive env-backup.tar.gz.

Extract it:

tar -xzvf env-backup.tar.gz

Now you'll have an env-backup/ folder containing:

venv-requirements.txt
system-site-packages.txt
environment-variables.txt

Install the packages:

pip install --no-cache-dir -r env-backup/venv-requirements.txt

[Optional but Recommended] If some packages were only available in the system (system-site-packages.txt) and are not yet installed (say OpenCV was preinstalled manually), install them using apt:

Example:

sudo apt install python3-opencv

Or if you installed OpenCV manually earlier from source, you might need to do it again.

---

✅ Step 5: Restore Environment Variables (if you had any)
If you had saved environment variables (environment-variables.txt), you can reload them:

set -o allexport
source env-backup/environment-variables.txt
set +o allexport

This will export all those environment variables into your current session.
(Or you can put them into ~/.bashrc if you want them permanently available.)

---

✅ Step 6: Verify Everything
You can check:

python --version
pip list

✅ Compare it to your backup's venv-requirements.txt and python-version.txt to ensure they match!

Optionally, you can install pipdeptree and check the dependency tree:

pip install pipdeptree
pipdeptree