How machine vision supports analyses of real-time hyperspectral image data
Find out how easy this can be by watching the video from our presentation at CHII 2018 in Graz:
Spectroscopy enables its user to identify spectral features that are not visible for common cameras or the human eye. Those features usually are directly related to the optical properties of the analysed surface. Due to the fact that each material has a different spectral signature such data has the potential not only to separate specific materials from others, but also allowing it to make qualitative statements on the analysed object. Spectral imaging in a next step allows to examine a spatial distribution of different materials and quality differences.
Hyperspectral snapshot cameras, such as the FireflEYE S185 from Cubert, provide a full data cube immediately – without the limitations of common spectral imaging systems based on line scanning. This gives the user the advantage to work on hyperspectral live image data, tracking processes where changes are hidden in spectral features.
Classification problem: Three different samples, in this case herbs, have quite similar spectral reflectance signals. Integrating machine learning solves this problem and allows to classify the samples correctly, even in the live data stream.
In order to support users in analysing their data Cubert, together with perClass BV, integrates a state-of-the-art classification solution based on machine learning to their Cubert Utils software. By this users can – within minutes – collect data, export it to the perClass Mira software, find complex classification models full automatically and apply them as plug-in to the live data stream in Cubert’s control software. Sophisticated knowledge on the background of machine learning and classification techniques is not necessary.
Some error has occured.