How much can you trust fragmentation measurement technologies and how laser-based technology can earn that trust
Joint Webinar: IOM3/AusIMM Metallurgical Society - How much can you trust fragmentation measurement technologies and how laser-based technology can earn that trust, presented by Dr Matthew Thurley
About this event
Fragmentation measurement is notoriously tricky to do in a repeatable and quantitative way with computer-vision methods. The goal of a fragmentation measurement system is to provide automated, consistent measurement for reliable feedback to blasting and process control in the plant.
However, highly variable image contrast due to varying lighting and material color significantly hinders the ability for image analysis to detect realistic particle boundaries and measure in a reliable way.
In addition, there are many sources of error common to all imaging-based methods, and photo-based methods are typically unable to address these adequately.
Participants will be introduced to the measurement challenges and critical sources of measurement error that affect and hinder computer-vision based fragmentation measurement methods.
Furthermore, examples from two different methods for fragmentation measurement based on laser technologies are presented that overcome these challenges and deliver the goal of automated reliable measurement. Measured data is presented from crusher output and SAG mill feed illustrating the impact of these sources of error and demonstrating the capacity to measure real and significant variations in size distribution that would otherwise be missed.
Speaker/s
Dr Matthew Thurley
He has a background in software engineering and computer vision, has worked at CSIRO with measurement technology for mining, and as Associate Professor in Image Analysis at Lulea University of Technology in Sweden.
He is an internationally leading researcher and developer of automated online fragmentation measurement solutions and is currently Principal Scientist at Innovative Machine Vision.
Date and Time
11.00am – 12.00am (UTC+08:00)