Nondestructive freshness discriminating of shrimp using visible/near-infrared hyperspectral imaging technique and deep learning algorithm

####Python codes of algorithms
1.SAEs-LR_train.py. #Train SAEs and LR by ramdon pixels.
2.SAEs-LR_classify.py. #Use trained SAEs-LR model for freshness classifiction (by mean spectra)
3.Classification_maps.py. #Use SAEs-LR model to create classification maps for hyperspectral images.

####data
folder sample_rois, #All pixels in ROI of shrimps (only 1 ROIs were uploaded because of the uploading time issue)
folder train_pixels, #The 102400 ramdon pixel spectra used to train the SAEs-LR model
file Meanspectra.csv, #The mean spectra of samples and the original experimantal data in this paper

####logs
LR_model #LR model struction file
LR_model.weights.hdf5 #The weights of trained LR model
SAEs_model #SAEs model struction file
SAEs_model.weights.hdfs #The weights of trained SAEs model
SAEs_model.train.log #The train error of SAEs model

####results
data, figures, and classification images created by the algorithms