35 results about "Prunella asiatica" patented technology

The invention discloses a hyperspectral method for identifying the producing area of selfheal on the basis of large model reprogramming (Model Reprogramming). The method comprises the following steps: (1) collecting hyperspectral data of selfheal samples from different producing areas; (2) correcting and preprocessing the collected hyperspectral data to obtain one-dimensional spica prunellae hyperspectral data; (3) designing an input adaptation function, and converting the one-dimensional spica prunellae hyperspectral data into a pseudo spectral signal meeting the input requirement of a large model; and (4) inputting the pseudo-spectral signal into a pre-trained large model, then mapping an original output category of the large model to a selfheal producing area category, and realizing accurate identification of the selfheal producing area by only training a reprogramming vector parameter in an input adaptation function. According to the method, a complex deep learning model does not need to be trained from zero, dependence on large-scale labeled hyperspectral data and consumption of computing resources are remarkably reduced, and a new normal form is provided for identifying the producing area of selfheal.