Back to Projects
ResearchAI-Based Industrial Inspection System

Smart Toolkit Inspector

Industrial computer vision system for automated toolkit inspection using depth sensing and real-time object detection.

96.9%

Segmentation

93.2%

Detection

CMC Journal

Published

2025

Year

Tech Stack

YOLOv11PythonOpenCVIntel RealSense SDKComputer VisionDepth EstimationRGB-D Imaging

What I Built

  • Developed an AI-powered toolkit inspection system for automated identification and dimensional analysis of industrial tools.
  • Used RGB-D imaging with Intel RealSense cameras to detect sockets and deep sockets in real time on an assembly line.
  • Combined object detection and edge-based measurement pipelines for accurate, repeatable industrial measurements.

Scalability & Engineering

  • Optimized inference workflows to reduce real-time latency for industrial environments with strict throughput requirements.
  • Designed lightweight processing pipelines to support edge deployment scenarios without cloud dependency.
  • Improved measurement precision using depth + RGB fusion techniques for reliable dimensional analysis.

Ownership & Impact

  • Led development of detection workflows and production-focused implementation under real industrial constraints.
  • Worked closely with industrial deployment requirements, ensuring reliability in noisy, high-throughput environments.
  • Resulted in a peer-reviewed publication in Computers, Materials & Continua (CMC) — an international journal.

Keywords

Computer VisionYOLOEdge AIIndustrial AutomationDepth ImagingReal-Time DetectionAI Systems