2T2D spectroscopy and deep learning for soil MPs quantification
Description
This dataset is primarily used for quantitative analysis of microplastic content in soil by combining DRSN-SpCA with 2T2D spectra. The code is written in Python, and the data is stored in .h5 format. main.py: This is the main function, which includes data import, model training, and result prediction. model_2T2D.py: The DRSN-SpCA model is encapsulated within it. early_stopping.py: This is the early stopping function. function_Function.py: The functions required by main.py are encapsulated within it. The specific data is available via DOI: 10.17632/zhdxvgzkfk.1 (PE_1), DOI: 10.17632/d2xz4sby82.1 (PE_2), DOI: 10.17632/2jf5gwrrg9.1 (PE_3_4), DOI: 10.17632/ksxndsxdgs.1 (PET_1), DOI: 10.17632/yp5kjgv58s.1 (PET_2), DOI: 10.17632/pybh5kgwz8.1 (PET_3_4), DOI: 10.17632/jdtnfgd5v7.1 (PP_1), DOI: 10.17632/spzk3vstdf.1 (PP_2), DOI: 10.17632/4wy2nrjbgc.1 (PP_3_4), DOI: 10.17632/wv8mn99ccb.1 (PS_1), DOI: 10.17632/4jy8k7j7ph.1 (PS_2), DOI: 10.17632/txrxrpks44.1 (PS_3_4).
Files
Steps to reproduce
First, the infrared (IR) spectra of soil-microplastics (MPs) mixtures were acquired using a Fourier transform infrared spectrometer. Then, the IR spectra were subjected to data augmentation, baseline correction, Savitzky-Golay filtering, standard normal variate and multiplicative scatter correction processing. Subsequently, they were transformed into two-trace two-dimensional (2T2D) correlation spectra. Finally, the combination of 2T2D spectroscopy (auto-synchronous, auto-asynchronous, sy-asynchronous, cross-synchronous and cross-asynchronous spectra) and a deep residual shrinkage network embedded with channel attention and strip attention (DRSN-SpCA) was employed for the quantitative analysis of MPs in soil.
Institutions
- Hefei University of TechnologyAnhui, Hefei