Research of using near infrared technology to quantitatively analyze amylose content of paddy rice single grain

The invention relates to crop single grain seed amylose content detection, and specifically discloses a paddy rice single grain seed amylose detection model construction method and applications thereof. The construction method comprises the following steps: collecting a diffuse reflection spectrum of a sample, selecting an effective wavelength, carrying out a smoothing treatment (9 point window), a first difference treatment, and a standardization treatment in sequence, utilizing partial least squares (PLS) to reduce dimensionality, adopting PLSR and a support vector machine (SVR) method to carry out regression related analysis on the spectrum matrix information and reference value matrix information through a chemometrics method, and establishing a crop single grain seed amylose detection model. In detection, the near infrared spectral data of an unknown sample is obtained under same conditions; the data is preprocessed by a same method, the characteristics of the data are extracted by a same technology, and