Moisture Determination for Fine-Sized Copper Ore by Computer Vision and Thermovision Methods

Area of Science:

• Materials Science
• Mineral Processing
• Computer Vision

Background:

• Bulk material moisture content significantly affects dry grinding efficiency, particle characteristics, and powder transport.
• Accurate on-site moisture estimation is crucial for optimizing industrial processes like dry grinding.

Purpose of the Study:

• To develop and evaluate computer vision and thermovision techniques for real-time copper ore moisture estimation.
• To investigate the impact of particle size on moisture estimation accuracy.

Main Methods:

• Utilized computer vision and thermovision imaging (standard and macro scales) on copper ore (0-2 mm particle size, 0.5-11% moisture).
• Developed and cross-validated models for moisture estimation based on image analysis.
• Analyzed the influence of particle size on estimation results.

Main Results:

• Median-intensity vision images showed a monotonic relationship with copper ore moisture content up to approximately 5%.
• Thermograms required sophisticated computer vision processing, not just mean temperature analysis, for effective moisture estimation.
• Particle size was found to influence the accuracy of moisture estimation.

Conclusions:

• Computer vision offers a viable method for on-site copper ore moisture estimation, particularly for lower moisture ranges.
• Advanced image processing is necessary for accurate moisture determination using thermovision data.
• The developed techniques can aid in optimizing dry grinding operations by providing real-time moisture feedback.