How to Classify Minerals Using Machine Learning

The ZEN AI Toolkit, inclulding ZEN Intellesis and ZEN Image Analysis modules, provides ZEISS microscopes with machine-learning image segmentation and classification using image analysis and object measurements.

The modules deliver automated minerals identification, classification, measurement, and quantification. They are available for use with ZEISS light microscopes, from the operator-driven Axiolab 5 to the fully automated Axioscan 7 Geo.

Automated Sample Analysis

Light microscopes are used on a daily basis for evaluating mining samples during the exploration of new resources and to assess the performance of the recovery process.

Operators must conduct a rigorous and precise evaluation of samples, a labor-intensive task that consumes a significant amount of time during the day.

Using ZEN Intellesis and ZEN Image Analysis, mineralogy classification and measurement are performed automatically, allowing the operator to focus on data interpretation and other vital tasks.

When combined with ZEN Image Analysis, ZEN Intellesis offers state-of-the-art machine learning, which is trainable on-site, for object classification and measurement. This helps to differentiate mineralogy, resulting in improved time to results and efficiency.

Reflected light image acquired with ZEISS Axiolab 5. Sample of mil led copper ore mounted in a resin block and previously prepared for SEM analysis. Remnants of a carbon coating remain on some surfaces.

Figure 1. Reflected light image acquired with ZEISS Axiolab 5. Sample of milled copper ore mounted in a resin block and previously prepared for SEM analysis. Remnants of a carbon coating remain on some surfaces. Image Credit: ZEISS Natural Resources

Artificial color map following Intellesis classification. Despite the challenges presented by sample preparation, e.g., the impossibility of obtaining a transmitted light image, surface scratches, and gauges, Intellesis automatically classifies the individual constituents of this image. Resin in light grey, gangue in dark grey, pyrite in brown, chalcopyrite in orange, and chalcocite in blue.

Figure 2. Artificial color map following Intellesis classification. Despite the challenges presented by sample preparation, e.g., the impossibility of obtaining a transmitted light image, surface scratches, and gauges, Intellesis automatically classifies the individual constituents of this image. Resin in light grey, gangue in dark grey, pyrite in brown, chalcopyrite in orange, and chalcocite in blue. Image Credit: ZEISS Natural Resources

Consistent Sample Evaluation

Quality monitoring is a continual process dependent on data being fed into an information loop from which operational decisions are made. This enables strategic planning using the evidence gathered over a period of time.

Reliable, consistent data is crucial. Unlike human operators, the ZEN Intellesis and Image Analysis modules do not suffer fatigue on long days and promise consistent results after each sample.

Following classification, the Image Analysis module automatic ally measures the individual constituents and provides statistical information, including modal distribution, size, area, perimeter, etc, as well as the modal proportion of phases present.

Figure 3. Following classification, the Image Analysis module automatically measures the individual constituents and provides statistical information, including modal distribution, size, area, perimeter, etc, as well as the modal proportion of phases present. Image Credit: ZEISS Natural Resources

Flexibility in Application

Flexibility and ease of use are key factors when using the ZEN modules. Machine learning models can be trained at the local level without needing to consult artificial intelligence (AI) experts. They can also be updated and refined over time, offering greater flexibility for materials classification.

Images from any ZEISS microscope can be processed, offering all levels of operation ranging from operator-driven image capture with Axiolab 5 to complete automation with Axioscan 7 Geo.

The wealth of information that can be leveraged using the fully automated Axioscan 7 Geo lends itself to a detailed and comprehensive analysis of the sample matter. Axioscan 7 Geo can capture images using various contrast methods and numerous angles of polarization, offering a high density of pixel information that can be evaluated and interpreted by artificial intelligence (AI) methods.

After implementation of Intellesis segmentation and image analysis, the results can be analyzed further with the application of the Petrography Analysis Toolbox (PetPAT). PetPAT is able to access the wealth of data the Axioscan 7 Geo images capture to provide outputs, including the angle of extinction relative to the analyzer, grain orientation, grain size distribution, the position of grain boundaries, and textural information.

How to Classify Minerals Using Machine Learning

Image Credit: ZEISS Natural Resources

This information has been sourced, reviewed and adapted from materials provided by ZEISS Natural Resources.

For more information on this source, please visit ZEISS Natural Resources.

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Carl Zeiss - Raw Materials. (2024, October 01). How to Classify Minerals Using Machine Learning. AZoMining. Retrieved on November 21, 2024 from https://www.azomining.com/Article.aspx?ArticleID=1729.

  • MLA

    Carl Zeiss - Raw Materials. "How to Classify Minerals Using Machine Learning". AZoMining. 21 November 2024. <https://www.azomining.com/Article.aspx?ArticleID=1729>.

  • Chicago

    Carl Zeiss - Raw Materials. "How to Classify Minerals Using Machine Learning". AZoMining. https://www.azomining.com/Article.aspx?ArticleID=1729. (accessed November 21, 2024).

  • Harvard

    Carl Zeiss - Raw Materials. 2024. How to Classify Minerals Using Machine Learning. AZoMining, viewed 21 November 2024, https://www.azomining.com/Article.aspx?ArticleID=1729.

Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.