Research On Individual Tree Location Technology Based On High-Resolution Image

With the advent of the era of big data,the rapid development of 3S technologies and computer image processing technology,the forest resource survey is moving from extensive stand(small class)level inspection to tree(individual trees)level precision survey.The ultimate in precision forestry is the accurate monitoring of individual-tree growth,and the establishment of a individual-tree database is the basis of this work,where individual-tree location technology is the key.This paper conducts an empirical study on the individual-tree location technology with different spatial distribution based on high-resolution images.Firstly,according to the study of related literatures on individual-tree extraction at home and abroad,the technical scheme based on individual-tree crown extraction and determination of individual-tree location information is determined.To this end,in·depth discussion of image preprocessing,crown extraction algorithm,individual trees location and individual-tree database establishment and related theory,methods and techniques based on Python+OpenCV experimental platform.Secondly,combined with the experimental data,the technical scheme of individual-tree location is mainly discussed:Thirdly,the woodland region is extracted by using the color index based on RGB color gamut;then the image processing steps such as grayscale transformation,threshold segmentation,corrosion expansion,and distance transformation are used to obtain the marker.The image is further extracted by the marker-based watershed algorithm;the relati-ve eoordinates of the individual trees are determined by calculating the center point of the crown profile;finally,the coordinates are converted into geographic coordinates,and a individual-tree database is established.Finally,for two types of individual trees(spread wood or four-sided trees in the city,forest trees in the forest),select the verifiable Beijing University of Forestry campusfs Beijing No.2 remote sensing image and the peak of the Jiu Feng forest park in unmanned aerial vehicle remote sensing image,the actual verification and improvement analysis of the above individual-tree location technology scheme,laid the foundation for further research.