Projects:2021s1-13007 Optical Character Recognition for Hazardous Area Equipment Labels
Devices installed in hazardous areas of a site typically have a label attached to them. This label is either engraved or etched with critical information such as location, tag number, certifications, model number and manufacturer of the device. Currently the technicians manually enter this data into a hazardous area inspection sheet (HAIS) using tablets and the equipment is then visually inspected and photographed. These inspections are performed on thousands of pieces of equipment at each site and reinspected once every 4 years, making it critical that the process be as efficient as possible.
Contents
Introduction
The manual inspection and input into HAIS using tablets is a highly tedious and time consuming task. GPA Engineering is seeking a way in which this information could be input in an semi-automated fashion using modern machine learning, computer vision and optical character recognition (OCR) techniques. There is additional scope for the research of lighting, capture and image manipulation techniques to enhance OCR capability in difficult environments or when equipment labels are either corroded or not mounted on a flat surface.
Project team
Project students
- Duncan Cameron
- Thomas Butler
Supervisors
- Dr. Danny Gibbins
- Dr. Hong Gunn Chu
- Mr. Josh Weber (GPA Engineering)
- Mr. Dave Johnson (GPA Engineering)
Objectives
Set of objectives
Background
Method
Results
Conclusion
References
[1]
[2]