画像範囲値を用いた同質領域判定手法によるシイ・カシ萌芽林とスギ・ヒノキ人工林の樹冠抽出

Translated title of the contribution: The Analysis of Forest Crowns by Same-Color-Region Determining Method Utilized Image Range Features of Castanopsis cuspidata Coppice Woodland and Cryptomeria japonica and Chamaecyparis obtusa Artificial Forest

朝廣 和夫, 井上 光平, 重松 敏則, 瀬戸島 政博

Research output: Contribution to journalArticle

Abstract

This research attempted to reveal that utilization of aerial photographs image range feature for classification of forest crowns by Same-Color-Region Determining Method of Castanopsis cuspidate coppice woodland and Cryptomeria japonica and Chamaecyparis obtusa artificial forest in northern part of Kyusyu, Japan. This research consist from two parts, one is the method improving applied by image range features such as standard deviation and cumulative frequency of tree crown edge for parameter setting. And the other is crown width classification accuracy was proved by field survey data of each tree crowns. As a results, supervised classifications of this improved method were capable of correctly classifying coppice woodland canopies with radii greater than 1.5 m (r<+-1.0 m) to an accuracy of 34.4%. For artificial forest, 41.4% of trees with tree crown radii greater than 1.0 m (r<+-0.5 m) could be correctly classified.
Original languageJapanese
Pages (from-to)593-598
Number of pages6
Journalランドスケープ研究 : 日本造園学会誌 : journal of the Japanese Institute of Landscape Architecture
Volume71
Issue number5
DOIs
Publication statusPublished - 2008

Fingerprint

coppice
woodland
image classification
aerial photograph
field survey
canopy
analysis
method

Cite this

@article{fcd66400071e4275ab7a91dd3f9ef459,
title = "画像範囲値を用いた同質領域判定手法によるシイ・カシ萌芽林とスギ・ヒノキ人工林の樹冠抽出",
abstract = "This research attempted to reveal that utilization of aerial photographs image range feature for classification of forest crowns by Same-Color-Region Determining Method of Castanopsis cuspidate coppice woodland and Cryptomeria japonica and Chamaecyparis obtusa artificial forest in northern part of Kyusyu, Japan. This research consist from two parts, one is the method improving applied by image range features such as standard deviation and cumulative frequency of tree crown edge for parameter setting. And the other is crown width classification accuracy was proved by field survey data of each tree crowns. As a results, supervised classifications of this improved method were capable of correctly classifying coppice woodland canopies with radii greater than 1.5 m (r<+-1.0 m) to an accuracy of 34.4{\%}. For artificial forest, 41.4{\%} of trees with tree crown radii greater than 1.0 m (r<+-0.5 m) could be correctly classified.",
author = "和夫 朝廣 and 光平 井上 and 敏則 重松 and 政博 瀬戸島",
year = "2008",
doi = "10.5632/jila.71.593",
language = "Japanese",
volume = "71",
pages = "593--598",
journal = "ランドスケープ研究 : 日本造園学会誌 : journal of the Japanese Institute of Landscape Architecture",
issn = "1340-8984",
publisher = "公益社団法人 日本造園学会",
number = "5",

}

TY - JOUR

T1 - 画像範囲値を用いた同質領域判定手法によるシイ・カシ萌芽林とスギ・ヒノキ人工林の樹冠抽出

AU - 朝廣, 和夫

AU - 井上, 光平

AU - 重松, 敏則

AU - 瀬戸島, 政博

PY - 2008

Y1 - 2008

N2 - This research attempted to reveal that utilization of aerial photographs image range feature for classification of forest crowns by Same-Color-Region Determining Method of Castanopsis cuspidate coppice woodland and Cryptomeria japonica and Chamaecyparis obtusa artificial forest in northern part of Kyusyu, Japan. This research consist from two parts, one is the method improving applied by image range features such as standard deviation and cumulative frequency of tree crown edge for parameter setting. And the other is crown width classification accuracy was proved by field survey data of each tree crowns. As a results, supervised classifications of this improved method were capable of correctly classifying coppice woodland canopies with radii greater than 1.5 m (r<+-1.0 m) to an accuracy of 34.4%. For artificial forest, 41.4% of trees with tree crown radii greater than 1.0 m (r<+-0.5 m) could be correctly classified.

AB - This research attempted to reveal that utilization of aerial photographs image range feature for classification of forest crowns by Same-Color-Region Determining Method of Castanopsis cuspidate coppice woodland and Cryptomeria japonica and Chamaecyparis obtusa artificial forest in northern part of Kyusyu, Japan. This research consist from two parts, one is the method improving applied by image range features such as standard deviation and cumulative frequency of tree crown edge for parameter setting. And the other is crown width classification accuracy was proved by field survey data of each tree crowns. As a results, supervised classifications of this improved method were capable of correctly classifying coppice woodland canopies with radii greater than 1.5 m (r<+-1.0 m) to an accuracy of 34.4%. For artificial forest, 41.4% of trees with tree crown radii greater than 1.0 m (r<+-0.5 m) could be correctly classified.

U2 - 10.5632/jila.71.593

DO - 10.5632/jila.71.593

M3 - 記事

VL - 71

SP - 593

EP - 598

JO - ランドスケープ研究 : 日本造園学会誌 : journal of the Japanese Institute of Landscape Architecture

JF - ランドスケープ研究 : 日本造園学会誌 : journal of the Japanese Institute of Landscape Architecture

SN - 1340-8984

IS - 5

ER -