Premium
Direct Quantification of Mixed Organic Acids Based on Spectral Image with Deep Learning
Author(s) -
Wang Wenjing,
Luo Run,
Duan Qiannan,
Feng Yunjin,
Chen Jiayuan,
Huang Yicai,
Bi Sifan,
Liu Fenli,
Lee Jianchao
Publication year - 2021
Publication title -
chemistryselect
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.437
H-Index - 34
ISSN - 2365-6549
DOI - 10.1002/slct.202100444
Subject(s) - image (mathematics) , artificial intelligence , deep learning , chemistry , computer science , environmental science , environmental chemistry , pattern recognition (psychology)
Identification of organic acid is crucial indicator for field of environment, healthcare and industry, which involves the time‐consuming and labor‐intensive progress. The simultaneous detection for complex mixture becomes a new target when large‐scale instruments is the mainstream measurement approach. With the burst development of machine learning in solving complex problem, the progress is being made for quick detection of mixtures. The next challenge is to provide sufficient data to satisfy the need of algorithm. In this work, we proposed the modified spectrometer technology combined with the deep learning to quantify the mixed organic acids. Organic acids interact with light in the various background to map the information in the form of images. Deep learning could establish the relationship between images and concentrations. According to the result, trained convolutional neural network could achieve the goal of simultaneous measurement of organic acids, which speeds up the implementation in the parallel detection and suggests new twists in the field of chemistry.