Hyperspectral imaging coupled with CNN: A powerful approach for quantitative identification of feather meal and fish by-product meal adulterated in marine fishmeal

•Three CNNs were developed for classification and regression of NIR-HSI data.•The proposed CNNs outperformed PLS and SVM on both classification and regression.•CNNs using < 20 effective wavelengths achieved results comparable to full wavelengths.•NIR-HSI with CNNs calibration can identify MFM adulterated with HFM, FBM, or HFM-FBM.