CN-121982431-A - Tree classification method and related equipment based on laser point cloud data and remote sensing images

CN121982431ACN 121982431 ACN121982431 ACN 121982431ACN-121982431-A

Abstract

The application provides a tree classification method and related equipment based on laser point cloud data and remote sensing images, wherein the method comprises the steps of acquiring the laser point cloud data of a target area and the remote sensing images of a long-time sequence; and inputting the structural features and the physical features into a pre-trained neural network model to classify tree species, thereby obtaining classification results. The tree classification method and the related equipment based on the laser point cloud data and the remote sensing image are simple and convenient, can effectively distinguish trees with similar morphological structures, and have high classification precision.

Inventors

JIANG MIAO
CHENG MINGHUA

Assignees

中国冶金地质总局矿产资源研究院

Dates

Publication Date: 20260505
Application Date: 20260403

Claims (10)

1. A tree species classification method based on laser point cloud data and remote sensing images is characterized by comprising the following steps: Acquiring laser point cloud data of a target area and a long-time sequence remote sensing image; extracting structural features of trees based on the laser point cloud data, and extracting weather features based on the remote sensing images; and inputting the structural features and the physical features into a pre-trained neural network model to classify tree species, and obtaining classification results.
2. The tree species classification method based on laser point cloud data and remote sensing image of claim 1 wherein the remote sensing image covers at least one complete growth cycle of the target area.
3. The tree species classification method based on laser point cloud data and remote sensing images of claim 1, wherein the structural features include tree height, breast diameter, crown area, crown volume, under-branch height, trunk volume and trunk tortuosity of the tree.
4. The method of claim 1, wherein the climatic features include a growth season beginning period, a growth season ending period, and a growth season length.
5. The tree classification method based on laser point cloud data and remote sensing images according to claim 1, wherein the inputting the structural features and the physical features into a pre-trained neural network model for tree classification comprises: And carrying out standardization processing on the structural features and the climatic features.
6. The tree species classification method based on laser point cloud data and remote sensing images according to claim 1, wherein the neural network model comprises an input layer, a first hidden layer, a second hidden layer and an output layer, the input layer comprises 11 neurons, the first hidden layer comprises 100 neurons, the second hidden layer comprises 50 neurons, and the number of neurons of the output layer is the number of tree species categories.
7. The tree species classification method based on laser point cloud data and remote sensing images according to claim 1, wherein an activation function adopted by the neural network model is a linear rectification function, the neural network model adopts an adaptive moment estimation optimizer to update parameters, the endurance coefficient of an early-stop mechanism of the neural network model is 20 rounds, and the maximum iteration number is 500 rounds.
8. The tree species classification method based on laser point cloud data and remote sensing images of claim 1, wherein the classification result comprises a tree species classification confusion matrix thermodynamic diagram and a tree species classification report, the tree species classification report comprising an accuracy rate, a recall rate, and an F1 score.
9. Tree species classifying device based on laser point cloud data and remote sensing image, characterized by comprising: the acquisition module is configured to acquire laser point cloud data of a target area and remote sensing images of a long-time sequence; the feature extraction module is configured to extract structural features of trees based on the laser point cloud data and extract weather features based on the remote sensing images; And the classification module is configured to input the structural features and the physical features into a pre-trained neural network model for classification, so as to obtain a tree classification result.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the laser point cloud data and remote sensing image based tree species classification method as claimed in any one of claims 1 to 8 when the computer program is executed.

Description

Tree classification method and related equipment based on laser point cloud data and remote sensing images Technical Field The application relates to the technical field of tree classification, in particular to a tree classification method and related equipment based on laser point cloud data and remote sensing images. Background The existing tree classification method mainly relies on manual field investigation, and has the problems of low efficiency, high cost, strong subjectivity and the like. The laser radar (LiDAR) technology can acquire laser point cloud data of trees, and provides a new technical means for automatic identification of tree species. The laser radar technology can penetrate through vegetation canopy to obtain three-dimensional space structure information of trees by actively transmitting laser pulses and receiving echoes, and has the advantages of high precision, all weather, non-contact and the like. The method is used for forest resource investigation, biodiversity monitoring, carbon sink metering, accurate forestry management and the like, can solve the problems of low efficiency, high cost, strong subjectivity and the like of the traditional manual investigation, and realizes automation and high-precision identification of tree species. However, some trees have similar morphological structures and are difficult to distinguish effectively by only relying on point cloud features. Therefore, a tree classification method with better precision is needed. Disclosure of Invention Accordingly, the present application is directed to a tree classification method and related equipment based on laser point cloud data and remote sensing images for solving the above technical problems. The application provides a tree species classification method based on laser point cloud data and remote sensing images, which comprises the steps of obtaining laser point cloud data of a target area and remote sensing images of a long-time sequence, extracting structural features of trees based on the laser point cloud data, extracting climatic features based on the remote sensing images, and inputting the structural features and the climatic features into a pre-trained neural network model to perform tree species classification to obtain classification results. Further, the remote sensing image covers at least one complete growth cycle of the target area. Further, the structural features include tree height, breast diameter, crown area, crown volume, under-branch height, trunk volume and trunk tortuosity of the tree. Further, the climatic features include a growing season beginning period, a growing season ending period, and a growing season length. Further, the step of inputting the structural features and the climatic features into a pre-trained neural network model for tree classification comprises the step of carrying out standardization processing on the structural features and the climatic features. Further, the neural network model comprises an input layer, a first hidden layer, a second hidden layer and an output layer, wherein the input layer comprises 11 neurons, the first hidden layer comprises 100 neurons, the second hidden layer comprises 50 neurons, and the number of the neurons of the output layer is the number of tree species. Further, an activation function adopted by the neural network model is a linear rectification function, the neural network model adopts an adaptive moment estimation optimizer to update parameters, the endurance coefficient of an early-stop mechanism of the neural network model is 20 rounds, and the maximum iteration number is 500 rounds. Further, the classification results include a tree classification confusion matrix thermodynamic diagram and a tree classification report including accuracy, recall, and F1 score. The application provides a tree species classification device based on laser point cloud data and remote sensing images, which comprises an acquisition module, a feature extraction module and a classification module, wherein the acquisition module is configured to acquire laser point cloud data of a target area and remote sensing images of a long-time sequence, the feature extraction module is configured to extract structural features of trees based on the laser point cloud data and extract candidate features based on the remote sensing images, and the classification module is configured to input the structural features and the candidate features into a pre-trained neural network model for classification, so that tree species classification results are obtained. In a third aspect of the present application, there is provided an electronic device, including a memory, a processor and a computer program stored in the memory and executable on the processor, where the processor implements the tree classification method based on laser point cloud data and remote sensing images according to the first aspect as described above when the computer program is executed. In a f