Skip to main content
Back to case studies
Case studyCompleted

Computer Vision Surveillance System

Built an end-to-end computer vision research system for real-time object detection, threat classification and alert generation.

PythonPyTorchYOLOv5/v8OpenCVFlaskNumPyLinux

Project summary

Final-year research project using YOLO, OpenCV and Flask for real-time object detection, threat classification and alerting.

Domain

computer vision

Project Type

final-year research

Problem

  • Urban surveillance workflows often rely on manual monitoring and delayed threat recognition.

Architecture

  • Problem-first framing so reviewers understand the operating context before implementation.
  • Implementation details are tied to the stack and workflows in the project record.
  • Sensitive client and company details are sanitized while preserving technical credibility.
  • Business value and measurable outcomes are highlighted where evidence exists.

Key Decisions

  • Use only completed work supported by the résumé or public repository evidence.
  • Keep project claims specific, factual, and proportionate to available proof.
  • Separate confidential client or company information from public technical summaries.
  • Prioritize implementation decisions and outcomes over decorative presentation.

Implementation

  • Built a YOLO-based real-time object detection system with Flask API, OpenCV stream processing and automated alerting.
  • Documented the stack, business context, and relevant operational outcomes.
  • Prepared a concise recruiter summary and safe public case-study presentation.

Quality Gates

  • Public records include the problem, solution, stack, business value, and evidence status.
  • Sensitive details remain intentionally summarized.
  • No unsupported performance, certification, or capability claim is included.

Results

  • Demonstrated functional automated threat recognition with a real-time inference and alert pipeline.
  • Demonstrates completed work rather than tutorial or study-plan output.
  • Supports software, backend, payments, business-systems, infrastructure, or computer-vision roles.

Future Improvements

  • Add screenshots or architecture diagrams where safe to publish.
  • Add demo videos for public-facing work.
  • Attach additional verified performance or business-impact metrics when available.

Public Evidence

  • Monitoring
  • Docs