Citrus huanglongbing detection method, device, equipment and medium

Through the multi-color spatial feature parameter detection method based on fluorescent images, combined with chi-square test and random forest classifier, the problem of indistinguishable citrus yellow dragon disease and tinogenic yellow disease is solved, and efficient and accurate detection is achieved, supporting orchards Management decision making.

CN119006904APending Publication Date: 2024-11-22SOUTH CHINA AGRICULTURAL UNIVERSITY
0 Cites 0 Cited by

Patent Information

Application Number
CN202411063491.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-05
Publication Date
2024-11-22

Smart Images

  • Figure CN119006904A_ABST
    Figure CN119006904A_ABST
Patent Text Reader

Abstract

The invention relates to a Citrus Huanglongbing detection method, apparatus and device, and a medium. The method comprises the following steps: obtaining a disease portrait corresponding to each citrus fruit; calling a preset citrus disease classification model to predict the classification probability of the disease labels with the disease portraits of the citrus fruits being the Huanglongbing type according to the plurality of feature parameters in the plurality of color spaces; calculating and determining feature contribution degrees of a plurality of feature parameters in a plurality of color spaces in the disease portraits corresponding to the citrus fruits to the classification probability, and determining contribution grades to which contribution grading numerical frequency bands hit by the feature contribution degrees of the plurality of feature parameters in the plurality of color spaces belong, constructing a sample data set by using the feature parameters reaching the target contribution level in the disease portrait; and predicting a disease label of a corresponding Huanglongbing type for the disease portrait corresponding to the to-be-detected citrus fruit by adopting a citrus disease classification model trained by the sample data set. According to the application, nondestructive and rapid detection of the citrus fruit huanglongbing can be realized.
Need to check novelty before this filing date? Find Prior Art