Hami melon pesticide residue quantitative analysis method based on combination of hyperspectral imaging and Regression GAN

The invention discloses a quantitative analysis method for pesticide residues of Hami melons by combining hyperspectral imaging with Regression GAN, and aims to realize quantitative spectral data enhancement of the pesticide residues of the Hami melons and improve the quantitative analysis precision of the pesticide residues of the Hami melons. The method comprises the following steps: firstly, collecting hyperspectral data of the pesticide residues of the Hami melons, determining a reference value of the pesticide residues of the Hami melons by using a liquid chromatography tandem mass spectrometer, and forming a quantitative spectral data set by the hyperspectral data and the reference value; preprocessing the spectral data by using a maximum and minimum normalization method, and dividing the preprocessed quantitative spectral data set into a training set and a test set; inputting the training set data into a Regression GAN model for training, and generating a new quantitative spectrum data set after networ