Forest wood resource survey method based on SLAM and image identification

Abstract

The invention provides a forest wood resource survey method based on SLAM and image identification. The method includes the following steps: S1, performing data collection on a forest sample plot to be surveyed, thereby obtaining image data and environmental data covering the whole sample plot scene; S2, utilizing an SLAM method to process the obtained image data, thereby generating a sample plotthree-dimensional model; S3, performing single-body processing on the three-dimensional model, and finding corresponding areas in the image to perform identification, thereby obtaining parameters of the three-dimensional space coordinates, height, diameter at breast height, categories and number of woods; and S4, importing collected sample plot scene raw data and processed data into a forest resource management unit to perform management. According to the method, forest data are obtained through a visual camera and a sensor, the SLAM method is utilized to build the forest scene three-dimensional model, identification and analysis of the forest data are completed through point cloud segmentation and image identification, field work time is reduced, the efficiency of forest resource survey is improved, and verifiability of a forest resource survey result is guaranteed at the same time.

Description

technical field [0001] The invention relates to a method for investigating forest resources, and belongs to the technical fields of forest investigation, computer vision and image processing. Background technique [0002] Forest resources are an important part of natural resources and an important material basis for national economic and social development. As forestry resources are over-cut and utilized by humans, resulting in shortage of timber and deterioration of the ecological environment, humans have to conduct forest resource surveys and planning and management. The survey of forest resources based on forest land, forest trees, animals and plants growing in the forest area and their environmental conditions provides reliable information for agricultural and forestry departments, timber users and those who formulate and implement national policies. [0003] Forest tree survey is the main content of forest resources survey. The survey factors of forest tree survey main...

AI Technical Summary

Problems solved by technology

[0004] The present invention provides a forest resource investigation method based on SLAM and image recognition, aiming to provide a forest resource investigation method for forest resource identification and automatic acquisition of investigation factors, and to solve the manual dependence in the current forest resource investigation process Large size, high cost, low survey efficiency, long period, and the accuracy of the survey results cannot be verified, so as to improve the efficiency and accuracy of forest resource surveys and reduce the cost of forest resource surveys

Images