Scientists from the Novosibirsk State Technical University NETI have developed a stand that automatically sorts diamonds by color with an accuracy of 98%. The technology is based on hyperspectral imaging and artificial intelligence. The solution can simplify and speed up one of the most complex stages of rough diamond processing.
The development is carried out at the Department of Optical Information Technologies, Faculty of Physics and Technology, NSTU NETI. Scientists are creating an automatic complex capable of recognizing diamond shades and classifying stones into color categories without human intervention.
The system uses hyperspectral imaging, a method in which each stone is analyzed in dozens of spectral ranges at once. This allows you to record the slightest differences in color that are difficult to notice during normal inspection. The received data is processed by a technical vision system with elements of machine learning.
The key part of the stand was the multichannel spectrometer “Kolibri-2”. It registers reflected from diamond light and transmits information to the software module, where artificial intelligence algorithms compare spectral characteristics and determine the color category.
According to the developers, the system is capable of sorting up to five stones per second and distributing them into ten color groups. This level of detail is especially important in the diamond industry, where even small differences in shade directly affect the cost of the raw material and the quality of the subsequent cut.
The university notes that manual sorting requires highly qualified specialists, takes a lot of time and remains subjective. Automation of this stage will improve the accuracy of the assessment, reduce the influence of the human factor and speed up the processing of batches diamonds.
The completion of the equipment setup is planned for the coming months, after which testing of the system will begin. In the future, the development can be adapted for implementation at enterprises in the diamond mining and diamond processing industries, where fast and accurate sorting methods are especially in demand.
Source: Komsomolskaya Pravda
Photo: Alexey Bulatov
