Bursaphelenchus xylophilus tree identification method and system and storage medium

Abstract

The invention relates to the technical field of plant detection, and discloses a pine wood nematode disease tree identification method and system and a storage medium, by identifying suspected pine wood nematode disease trees and ground feature types in images, reverse masking is performed on the suspected pine wood nematode disease trees coinciding with ground features on which pine wood nematode cannot propagate, so that pine wood nematode disease identification is realized. According to the method, the suspected pine wood nematode disease tree which is overlapped with the ground feature which cannot be propagated by the pine wood nematode can be determined not to be the suspected pine wood nematode disease tree actually, the probability that other ground features are mistakenly predicted as the suspected pine wood nematode disease tree is reduced, and the identification precision of the pine wood nematode disease tree is improved.

Description

technical field [0001] The invention relates to the technical field of plant detection, in particular to a method, system and storage medium for identifying trees with pine wood nematode disease. Background technique [0002] Pine wood nematode (Bursaphelenchus xylophilus) is a disease and insect pest caused by the pine wood nematode, which can spread rapidly in forests and cause extensive infection and death of trees. Pine trees are the main target of infection. It is one of the most dangerous forest biological disasters in China. In 1982— In the 35 years in 2017, the disease caused more than 50 million pine trees to die, and the economic loss reached hundreds of billions of yuan, causing heavy losses to China's forestry ecology and economy. The outside of the diseased tree shows the visible features of discoloration and browning of the needles and wilting and drooping due to water loss. The traditional monitoring of pine wood nematode disease is mainly to monitor the disco...

AI Technical Summary

Problems solved by technology

However, when making label pictures in this invention, only the pictures containing the dead wood of pine wood nematode disease are marked, and the pictures without the dead wood of pine wood nematode disease are all marked as the background class, which will cause the characteristics of the background class pictures to be inconspicuous and semantic It is difficult for the segmentation model to correctly identify regions that do not contain pine wood nematode disease trees as background classes, and the accuracy of identifying pine wood nematode disease trees is low

Images