Development of a methodology to extract individual trees in a Japanese Cider plantation using LiDAR data

Abstract

       We developed a vector based extraction methodology of an individual tree in a Japanese Cider plantation. The developed methodology can extract each tree based on tree crown shape without the need of rasterization. Moreover, in order to extract trees with various crown sizes, the methodology consisted of two modules. One was for trees with an isolated tree crown shape. The other was for trees with suppressed tree crown shape. As a result of comparison between the geographical position of extracted tree and of measured tree, the accuracy of positioning was found to be within 1m. [This abstract is not included in the PDF]