Visualization and accuracy improvement of soil classification using laser-induced breakdown spectroscopy with deep learning

Deep learning method is applied to spectral detection due to the advantage of not needing feature engineering. In this work, the deep neural network (DNN) model is designed to perform data mining on the laser-induced breakdown spectroscopy (LIBS) spectra of the ore. The potential of heat diffusion for an affinity-based tran-sition embedding model is first used to perform nonlinear mapping of fully con-nected layer data in the DNN model. Compared with traditional methods, the DNN model has the highest recognition accuracy rate (75.92%). A training set up-date method based on DNN output is proposed, and the final model has a recog-nition accuracy of 85.54%. The method of training set update proposed in this work can not only obtain the sample labels quickly but also improve the accuracy of deep learning models. The results demonstrate that LIBS combined with the DNN model is a valuable tool for ore classification at a high accuracy rate.

Automated summary

Deep learning method is applied to spectral detection without feature engineering. A DNN model is designed to mine data from LIBS spectra of ore. Compared with traditional methods, the DNN model achieves the highest recognition accuracy rate (75.92%). A training set update method based on DNN output is proposed, resulting in a recognition accuracy of 85.54%. The combination of LIBS and DNN model proves to be a valuable tool for accurate ore classification.