PhD Researcher · Machine Vision for Textile Quality
Md Rasel Khandaker
I build low-cost vision systems that measure fabric colour objectively, so that small and medium textile factories can check shade consistency without an expensive laboratory instrument.
Same colour to the eye?
Five dyed samples beside their reference. Side by side they look alike, but at a CIEDE2000 limit of 2.0 two of them fail. My system makes that call from a webcam image in real time.
Colour differences computed with CIEDE2000 from the sRGB values shown; limit ΔE00 = 2.0.
Low-cost real-time fabric colour variation detection
A Python desktop application with a consumer USB webcam and daylight-balanced LED panels, instead of a spectrophotometer. The demo button opens a browser version of the same method that you can try with your own camera.
- Colour difference in CIEDE2000, with ΔE*ab and CMC(2:1) recorded
- Five readiness checks before a test: camera, exposure, stability and even lighting
- Pass/fail limit derived from the system's own measured noise
- Reports with per-frame CSV export for analysis
Publications
- Adoption of IoT-Enabled Quality Monitoring System in a Smart Factory M. R. Khandaker and S. P. A/P Thiagarajah · 6th International Conference on Technology and Innovation Management (ICTIM 2026), Multimedia University, Malaysia Presented
- A Low-Cost Machine Vision System for Real-Time Fabric Colour Variation Detection M. R. Khandaker and S. P. A/P Thiagarajah · targeting an IEEE venue In preparation
Open to research collaboration
Textile quality, colour measurement, machine vision and IoT in manufacturing.