AUTHOR=Li Jie-Qing , Wang Yuan-Zhong , Liu Hong-Gao 

TITLE=Application of spectral image processing with different dimensions combined with large-screen visualization in the identification of boletes species

JOURNAL=Frontiers in Microbiology

VOLUME=Volume 13 - 2022

YEAR=2023

URL=https://www.frontiersin.org/journals/microbiology/articles/10.3389/fmicb.2022.1036527

DOI=10.3389/fmicb.2022.1036527

ISSN=1664-302X

ABSTRACT=Boletes are favored by consumers because of their unique flavor, rich nutrition and delicious taste. However, the different nutritional values of each species of boletes lead to obvious price differences, so shoddy products appear on the market, which affects food safety. Therefore, it is very important to identify species of boletes. In this paper, two-dimensional correlation spectroscopy (2DCOS) images are generated by computer method, and three-dimensional correlation spectroscopy (3DCOS ) images are proposed and developed for the first time. Then, the original Mid-Infrared (MIR) spectroscopy data is adopted for support vector machine (SVM) modeling. Seven data sets are used to carry out Alexnet and residual network (Resnet) modeling, so we used 1707 infrared spectral data and 11949 spectral images to established 15 models for the identification of boletes species. The results show that the Resnet model is better than Alexnet and SVM. The SVM method needs to process complex feature data, the time cost is more than 11 times of the average time of other models, and the accuracy is not high enough, so it is not recommended to be used in data processing with large sample size. From the perspective of datasets, synchronous 2DCOS and Synchronous 3DCOS have the best modeling results, while one-dimensional (1D) MIR Spectrum dataset has the worst modeling results. After comprehensive analysis, the modeling effect of Resnet on the synchronous 2DCOS dataset is the best. When the accuracy is 100%, its loss value and time cost are the lowest. Finally, we use large-screen visualization technology to visually display the sample information of this research and obtain their distribution rules in terms of species and geographical location. This research shows that deep learning combined with 2DCOS and 3DCOS spectral image processing can effectively and accurately identify the species of boletes, and subsequent studies can try to apply it in other discrimination fields.