06Published2024

Retail Theft Detection

Real-time computer vision system for loss prevention

Researcher

Published in IJSREM — International Journal of Scientific Research in Engineering and Management
YOLOv8
Detection
DeepSORT
Tracking
IJSREM
Published
2024
Year

Detection Pipeline

01
Frame Capture
RTSP stream or video file ingested at 25fps via OpenCV.
02
Person Detection
YOLOv8 runs inference on each frame, detecting persons with bounding boxes.
03
Multi-Object Tracking
DeepSORT assigns persistent IDs to each detected person across frames.
04
Behavior Analysis
Trajectory and dwell-time analysis flags suspicious patterns (loitering, concealment zones).
05
Alert Generation
Flagged events logged with timestamps, track IDs, and annotated frame snapshots.

Background

Retail theft costs businesses billions annually. Most loss prevention relies on reactive CCTV review after an incident. This research explores a proactive, automated approach using deep learning for real-time detection and tracking.

Technical Approach

YOLOv8 was chosen for its balance of speed and accuracy — it runs at real-time fps on consumer-grade hardware. DeepSORT extends single-frame detection to full trajectory tracking by combining Kalman filter predictions with deep appearance features (Re-ID). The behavior analysis layer uses heuristic rules derived from LP domain knowledge: dwell time in high-theft zones, occlusion events near shelves, and unusual trajectory patterns.

Results

The system achieved high detection accuracy on the evaluation dataset. False positive rate was kept low through a confidence threshold tuned to minimize unnecessary alerts. The paper was accepted and published in the IJSREM peer-reviewed journal — my first academic publication.

Learnings

This project taught me the gap between academic benchmarks and real-world deployment. Lighting variation, camera angle, and occlusion are far harder in-the-wild than in curated datasets. The next iteration would incorporate domain adaptation and a confidence calibration layer.

Tech Stack

PythonYOLOv8DeepSORTOpenCVTensorFlowComputer VisionNumPy