Type identification method and volume measurement method of standing trees

Abstract

The invention discloses a type identification method and a volume measurement method of a living standing tree, and relates to the technical field of forest resource investigation. The species identification method includes: acquiring RGB-D images of different angles of the tested living wood; based on the RGB-D images of different angles, splicing to obtain a complete image of the tested living wood, and separating from the complete image of the tested living wood. Trunks, branches and leaves of the tested living standing trees; based on the trained deep convolutional neural network model, the separated tree trunks and leaves are identified, and the category with the highest confidence after weighted fusion is the tested living standing tree species. The embodiment of the present invention uses the deep convolutional neural network model to identify tree species to obtain tree species information, and then obtains the tree species structure model of the living standing tree to be tested, and calculates the volume in combination with the geometric principle, thereby improving the type identification accuracy of the living standing tree and the volume measurement accuracy. .

Description

technical field [0001] The invention relates to the technical field of forest resources investigation, in particular to a type identification method and a volume measurement method of living standing trees. Background technique [0002] The implementation of forest resources survey is an important way to obtain basic forest information. The species and volume of living trees are two important contents in forest resources survey. The traditional tree species identification method is mainly through observation, comparison and analysis of the main characteristics of each tree species, and then the tree species is manually identified step by step from macroscopic features to microscopic features. Volume refers to the volume of the bark on the trunk of a tree. For a long time, forestry workers have been using felled standard logs to obtain modeling samples for compiling forest numbers. Although this method has high precision, it requires a large number of trees to be felled. Thi...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the problem in the related art that the type identification and volume measurement accuracy of the measured standing trees are not high under the condition of tree canopy occlusion, the embodiment of the present invention provides a type identification method and a volume measurement method of living trees, using deep convolutional neural network The network model identifies tree species to obtain tree species information, and then obtains the tree species structure model of the measured living trees, and calculates the volume in combination with geometric principles, thereby improving the accuracy of species identification and volume measurement accuracy of living trees

Images