technology
About OreScanner
OreScanner is focused on delivering modern sensor solutions for the mining industry, providing superior decision-making capabilities at every stage of mineral processing. Our easy-to-install system provides real-time ore assessment above a conveyor belt or at the loading/unloading stage.
Technology Used
OreScanner employs a sensor fusion technique to obtain crucial information about your ore type. This approach involves combining multiple sensors to construct the necessary classification models and address mineral complexity and measurement environment challenges. We rely solely on non-destructive measurement techniques, and there is no need for ore sample preparation for our assessments. Our measurement principles are both environmentally friendly and safe, with no radiation emission involved, unlike traditional methods like PGNAA and X-Ray.
Sensors:
Computer Vision
Computer vision plays a vital role in modern ore sorting processes. Through the use of advanced imaging and machine learning algorithms, computer vision technology enables efficient and accurate identification of different types of ore based on their visual characteristics.
Recent advancements in hardware and sensor technology have led to the availability of cost-effective high-resolution cameras and imaging systems. These cameras, coupled with efficient image processing capabilities, enable accurate and detailed analysis of ore particles without requiring expensive or specialized equipment.
This technology revolutionizes traditional ore processing methods, enabling mining companies to optimize their operations and extract maximum value from their ore resources.
Multispectral imaging
The technique captures and analyzes the electromagnetic spectrum at different wavelengths. It involves capturing images of ore in multiple spectral bands, enabling the identification of materials based on their spectral signature.
Multispectral imaging can serve as a cost-effective and efficient technique for identifying and separating valuable minerals from a waste rock on a conveyor belt. Typically, multispectral imaging systems employ a combination of visible and infrared wavelengths to capture images of the ore stream as it passes through the sorting system.
Near-infrared (NIR) sensors
Near-infrared (NIR) sensors function by illuminating minerals with a near-infrared light source and measuring the resulting reflected light. As different minerals possess distinct spectral signatures within the NIR range, they can be identified and separated based on their unique composition.
Mid-infrared (Mid-IR) spectrometry
Mid-IR spectroscopy targets the molecular vibrational modes of chemical compounds, which provide detailed information about the chemical structure and composition of substances. This spectral region corresponds to fundamental vibrational frequencies and overtones, allowing for more specific identification and quantification of chemical bonds and functional groups.
Sonic sensors
Sonic sensors operate by emitting sound waves at a designated frequency and measuring the resulting vibrations in minerals. Each mineral possesses distinctive acoustic signatures, enabling their identification and separation based on their composition.
Data Processing
To accurately differentiate ore quality from waste or valueless rock, data from multiple sensors is analyzed using computer vision, machine learning (ML) and chemometric algorithms.
Why OreScanner
Most sorting and ore measurement technology suppliers available in the market only focus on one type of sensor, and there is a lack of cooperation between competing parties. As a result, clients are responsible for combining different technologies to obtain a more comprehensive service. However, it is difficult to achieve this without sufficient domain knowledge of each measurement technology.
In contrast, OreScanner is not tied to any particular technology and selects an array of sensors suitable for the specific application. Moreover, OreScanners are equipped with only the necessary sensors to provide the required data for confident decision-making in process control. This approach reduces the overall cost of sorting applications compared to other vendors who provide excessive data that is not required for performing the task.