Missing Person Identification Using AI

Abstract

This paper presents MissingPersonAI Pro, a web-based artificial intelligence system designed to support the identification of missing individuals through automated facial image comparison. Conventional missing-person searches often involve manual photograph examination, eyewitness information, and records maintained across separate sources, which can make the identification process slow and difficult to manage at scale. The proposed system addresses this challenge by combining facial recognition, image processing, similarity matching, and structured database management within a single platform. Administrators can register missing individuals by entering relevant personal information and uploading multiple facial photographs. These images are processed to obtain facial feature representations that can be stored for subsequent comparison. When a user submits an image, the system detects the face, extracts its features, and compares them with the stored representations to determine a potential matching record. The resulting match is displayed together with a similarity score and relevant information. The system also maintains identification history and provides PDF report generation to support documentation and further investigation. Python and Flask are used for application development, while SQLite manages user records, missing-person information, facial-image references, and matching history. Role-based access is incorporated to control administrative and general-user operations. The functional tests documented in the project report successfully covered authentication, person registration, image uploading, face detection, feature extraction, matching, and report generation. The proposed framework can reduce repetitive manual comparison and provide a structured technological aid for missing-person investigations. Future improvements may include real-time CCTV analysis, age-progression support, multi-face detection, mobile integration, and connection with authorized large-scale databases.

Introduction

The text presents MissingPersonAI Pro, an AI-powered web application designed to assist in identifying missing persons through facial recognition, similarity matching, database management, and automated reporting. The system aims to reduce the time and effort required for traditional manual comparison of photographs and separate record searches.

Problem

Traditional missing-person identification depends heavily on manual examination of photographs, eyewitness information, and scattered records. This process becomes difficult when large numbers of images must be examined and can be affected by:

• Differences in lighting and image quality

• Facial expressions and viewing angles

• Changes in appearance over time

• Large numbers of photographs and records

• Poor organization of previous identification attempts

These limitations create a need for an automated and structured identification system.

Objectives

The main objectives of MissingPersonAI Pro are to:

• Detect faces in uploaded images.

• Extract facial features and generate facial embeddings.

• Compare uploaded faces with registered missing-person records.

• Provide potential matches with a similarity/confidence score.

• Allow administrators to register missing persons and multiple facial images.

• Maintain identification and match history.

• Generate downloadable PDF reports containing match information and visual evidence.

• Provide authentication and role-based access to protect sensitive records.

Methodology

The system follows a sequence of steps:

1. Registration: Administrators enter missing-person details and upload multiple photographs.

2. Face processing: The system detects faces and converts them into numerical facial embeddings.

3. Database storage: Personal information, image references, and embeddings are stored in a structured database.

4. Image search: A user uploads an image through the web application.

5. Matching: The system detects the face, generates its embedding, and compares it with stored embeddings using similarity measures.

6. Result generation: The closest potential match is displayed with a confidence score.

7. Documentation: The identification attempt is stored in match history and a PDF report is generated.

8. Security: Authentication and role-based authorization control access to administrative functions.

Literature review

Previous research has explored numerous AI and computer-vision techniques for missing-person identification, including YOLOv8, CNNs, FaceNet, VGG-Face, VGG16, SVM, Haar Cascade, DeepFace, PCA, LBP, cosine similarity, and Euclidean distance.

Research has also investigated real-time CCTV recognition, mobile applications, geolocation, age progression, multi-source information, and automated alerts. However, practical challenges remain, particularly lighting and image-quality variations, aging, occlusion, changes in facial appearance, scalability, and privacy/security of facial data.

Existing vs. proposed system

Existing systems primarily depend on manual photograph comparison and basic database searches. They may lack automated facial matching, centralized records, match-history management, and automated reporting.

MissingPersonAI Pro integrates these functions into one web-based platform. It combines:

• Facial detection and recognition

• Facial embeddings

• Similarity-based matching

• Centralized database management

• Authentication and role-based access

• Match-history tracking

• Confidence scoring

• PDF report generation

Feasibility

The system is considered feasible in four major areas:

• Economic: Uses open-source technologies such as Python, Flask, and SQLite, reducing development costs.

• Operational: Provides interfaces for administrators and general users and automates repetitive identification tasks.

• Technical: Uses established technologies for web development, computer vision, databases, and facial recognition.

• Legal/Ethical: Recognizes the importance of authentication, access control, privacy, and responsible handling of facial and personal information.

Technologies used

The major technologies include:

• Python – backend and AI/image-processing implementation.

• Flask – web application framework.

• Computer vision and facial recognition models – face detection and embedding generation.

• SQLite – storage of user, missing-person, image, and matching information.

• HTML, CSS, JavaScript – web interface. facial feature extraction, similarity matching, structured data storage, authentication, match-history management, and PDF reporting in a single platform. The system is designed to reduce manual effort and

• FPDF – automated PDF report generation.

Conclusion

MissingPersonAI Pro demonstrates a web-based approach for supporting missing-person identification through artificial intelligence and facial-recognition techniques. The system enables administrators to register missing-person details and associated facial images, while authorized users can upload an image for identification. The submitted image is processed through face detection and facial feature extraction, and the resulting facial representation is compared with stored records to produce a matching result and confidence score. The system further integrates role-based access control, structured database management, match-history storage, and PDF report generation, providing a unified workflow for identification and documentation. The implemented system was evaluated through functional testing of authentication, authorization, missing-person registration, image processing, feature extraction, face matching, database operations, and report generation, with the documented test cases showing successful results for the tested scenarios. Overall, MissingPersonAI Pro provides a practical technological framework for reducing manual effort in missing-person identification while maintaining organized records and supporting documentation through generated reports.

References

[1] S. Schroff, D. Kalenichenko, and J. Philbin, “FaceNet: A Unified Embedding for Face Recognition and Clustering,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015. [2] O. M. Parkhi, A. Vedaldi, and A. Zisserman, “Deep Face Recognition,” British Machine Vision Conference (BMVC), 2015. [3] Y. Taigman, M. Yang, M. Ranzato, and L. Wolf, “DeepFace: Closing the Gap to Human-Level Performance in Face Verification,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2014. [4] F. Schroff, D. Kalenichenko, and J. Philbin, “Face Recognition Using Deep Neural Networks and Facial Embeddings,” IEEE Conference on Computer Vision and Pattern Recognition, 2015. [5] P. Viola and M. Jones, “Rapid Object Detection Using a Boosted Cascade of Simple Features,” Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2001. [6] G. Bradski, “The OpenCV Library,” Dr. Dobb’s Journal of Software Tools, 2000. [7] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet Classification with Deep Convolutional Neural Networks,” Advances in Neural Information Processing Systems, 2012. [8] K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016. [9] C. Szegedy et al., “Going Deeper with Convolutions,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015. [10] D. Patel, A. Solanki, J. Magajwala, and H. Sutariya, “AI-Based Missing Person Detection System,” 2025. VISHNU_3VY24MC077_Report(1)(1) [11] L. Sridhara Rao, A. Thanmai, M. Hemanth, and N. Balraj, “Missing Person Identification System Using Deep Learning,” 2025. VISHNU_3VY24MC077_Report(1)(1) [12] S. B. Patil, R. Fegade, P. Nanaware, and S. Kadu, “Finding Missing Person Using AI and ML,” 2025.

Copyright

Copyright © 2026 Vishnu ., Shruthi Rampure. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.