High canopy density stand crown area acquiring method based on images acquired by unmanned aerial vehicle

An acquisition method, UAV technology, applied in image analysis, image enhancement, image data processing, etc., can solve the problem of inability to accurately extract single tree canopy information and area, and achieve the effect of avoiding tedious work and improving work efficiency

Active Publication Date: 2017-12-08
NORTHEAST FORESTRY UNIVERSITY
View PDF5 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problem that the image data acquired by the existing UAVs cannot be accurately distinguished due to the mutual occlusion between tree crowns, resulting in the inability to accurately extract the information and area of ​​single tree crowns. Acquisition Method of Canopy Area of ​​High Canopy Density Stand Based on Man-machine Image

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • High canopy density stand crown area acquiring method based on images acquired by unmanned aerial vehicle
  • High canopy density stand crown area acquiring method based on images acquired by unmanned aerial vehicle
  • High canopy density stand crown area acquiring method based on images acquired by unmanned aerial vehicle

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0023] Specific embodiment one: a kind of high canopy density forest canopy area acquisition method based on unmanned aerial vehicle image comprises the following steps:

[0024] Step 1: Use drones to collect images of forest land, process the images to generate digital orthophoto maps (Digital Orthophoto Map, DOM), digital surface models (digital surface models, DSM) and digital elevation models (Digital elevation models, DEM) );

[0025] Step 2: The digital orthophoto image obtained in step 1 is processed by excess green (EXG) to obtain the forest area image, and the forest area image is processed by Sobel algorithm after binarization and image morphology processing. The sub detects the edge of the woodland;

[0026] Step 3: Perform median filtering on the digital surface model obtained in step 1 to obtain the pixel curves of each row and each column, and take out the minimum value points on the curves to obtain the boundaries between the tree crowns that block each other; ...

specific Embodiment approach 2

[0029] Specific embodiment 2: The difference between this embodiment and specific embodiment 1 is that in the step 1, drones are used to collect images of forest land, and the images are processed to generate digital orthophoto images, digital surface models and digital elevation models. The specific process is:

[0030] Step 11: Import the UAV images into the photogrammetry and modeling software Agisoft photoscan;

[0031] Step 1 and 2: According to the latest multi-view Figure three The three-dimensional reconstruction technology automatically calculates the position and attitude of the photo, and the internal orientation, relative orientation and absolute orientation are all automatically completed. The three-dimensional dense point cloud data with coordinate information is extracted from the original photo data taken by the drone without manual work. Additional intervention. The basic data required are images, pos data and control point data.

[0032] Step 13: Accordin...

specific Embodiment approach 3

[0037] Specific embodiment three: the difference between this embodiment and specific embodiment one or two is that in the step two, the digital orthophoto image obtained in the step one is processed with a super green feature algorithm to obtain a forest area image, and the forest area image is passed through The specific process of using the Sobel operator to detect the edge of the woodland after binarization and image morphology processing is as follows:

[0038] Step 21: According to the xcess green (EXG) feature index, the digital orthophoto image obtained in step 1 is grayscaled. The formula of the xcess green feature index is:

[0039] EXG=2ρ green -ρ red -ρ blue

[0040] Wherein said EXG is calculated to obtain the super green characteristic index value of the pixel point, ρ green ,ρ red ,ρ blue Respectively represent the reflection values ​​of the three bands of green, red and blue;

[0041] Step 22: Binarize the grayscale image obtained after being processed b...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a high canopy density stand crown area acquiring method based on images acquired by an unmanned aerial vehicle. The invention relates to high canopy density stand crown area acquiring methods. The invention aims at addressing the failure of current unmanned aerial vehicles to accurately extract crown information and areas of single trees because it is difficult to accurately differentiate crowns that are joined or cover one another from image data which is acquired by current unmanned aerial vehicles. According to the invention, the method herein includes the following steps: 1. generating a digital orthographic image, a digital surface model and a digital elevation model; 2. processing the digital orthographic image to obtain a forest region image which undergoes binary processing and image morphology processing, then using Sobel operator to detect forest edges; 3. conducting median filter on the digital surface model to obtain the pixel curves of respective rows and respective columns, picking out minimum values on the curves to obtain the boundaries among the crowns that cover one another; and 4. Combining the forest edges and the boundaries of the crowns which cover one another, using theHough transformation algorithm to detect circles and extract crown areas. According to the invention, the method is applied to remote sensing in forestry.

Description

technical field [0001] The invention relates to a method for acquiring the canopy area of ​​a stand with high canopy density. Background technique [0002] With the continuous development of drone technology, its application in forestry is becoming more and more extensive. As an important way to obtain high-resolution images, UAV aerial photography has the characteristics of low cost, high efficiency, and strong timeliness. It has gradually become a new way for forest resource investigation and monitoring. As a new type of technical equipment for obtaining data sources, light and small drones make up for the low resolution of traditional satellite remote sensing images and are easily affected by clouds with their advantages of low cost, lightness, flexibility, convenience, safety, and fast acquisition of high-resolution images. Insufficiencies such as data quality and reentry cycle limitations. At the same time, UAV aerial photogrammetry technology can realize the data col...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06T5/30G06T7/11G06T7/13G06T7/136G06T7/62
CPCG06T5/30G06T7/11G06T7/13G06T7/136G06T7/62
Inventor 林文树李祥吴金卓
Owner NORTHEAST FORESTRY UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap