AI-Powered Missing Person Identification System Using Computer Vision

Shruti Rowlo — Sun, 11 Oct 2026 14:39:25 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

I developed an AI-powered Missing Person Identification System that explores how computer vision can help compare facial features and support missing-person search efforts.

The project uses MediaPipe Face Landmarker to extract facial landmarks and a K-Nearest Neighbors (KNN) classifier to compare feature representations. The goal is to explore how machine learning can assist in identifying potential matches.

This is a prototype, not a verified identification service. Any potential match requires human review and independent verification.

GitHub: https://github.com/Shrutirowlo/missing_person_AI

Demo

Source code: https://github.com/Shrutirowlo/missing_person_AI

The repository contains the project implementation. Please refer to the README for setup instructions and available demonstrations.

Code

GitHub repository: https://github.com/Shrutirowlo/missing_person_AI

The repository contains the source code for the missing-person identification prototype. Please refer to the repository for implementation details and setup instructions.

How I Built It

The system explores a computer vision pipeline using MediaPipe Face Landmarker for facial landmark extraction and K-Nearest Neighbors (KNN) for classification.

The pipeline extracts facial landmark features from images, prepares feature vectors for training, and evaluates the classifier on held-out samples.

The project also explores traditional face-recognition approaches, including Fisherfaces and LBPH.

The implementation uses Python and machine-learning tools. See the repository for the exact dependencies and implementation details.

Why Does Open Innovation Matter?

Open-source computer vision and machine-learning tools make it possible for developers and researchers to inspect implementations, experiment with different approaches, and learn without depending entirely on proprietary APIs.

For a missing-person identification prototype, local processing can also help reduce the need to upload sensitive facial images to third-party AI services.