Method and system for identifying diseases and pests of brassica oleracea var. botrytis linnaeus based on hyperspectral imaging

Abstract

The invention discloses a method and system for identifying diseases and pests of Brassica oleracea var. botrytis Linnaeus using hyperspectral imaging. A crop growth model is first established and combined with current cultivation and meteorological data to predict disease and pest occurrence periods. When the predicted period is reached, visible and hyperspectral images of the plants are collected. The plant region is separated from the background to generate spectral images. Disease-sensitive spectral bands are extracted based on reflectance differences across bands, and corresponding disease or pest areas are located in the images. These areas are then input into a target disease and pest identification network model, which extracts features and accurately determines the specific type of disease or pest affecting Brassica oleracea var. botrytis.