草业学报2011，Vol.20Issue(2)：156-162,7.
基于分形维数的放牧地秃斑特征提取方法及其应用
Features and application of a rangeland detection algorithm based on fractal dimensions to detect characteristics of barren patch
摘要
Abstract
The area and circumference of barren patch in grazing grassland was calculated by GIS software (ArcView GIS 3.2a) using field digital imaging to measure the shape index and fractal dimensions of barren patches.This method showed that barren patches in grazing grassland of the degraded alpine meadows has typical fractal characteristics with shape index and fractal dimensions of 1.8 and 1.1 respectively. There were 120 least insert points following the typical barren patch boundary in degraded alpine grazing meadow and the relative shape index change rate was ＜5％ while the relative fractal dimension change rate was ＜0.5％. The method gives fast, accurate measurement, fine practicability and maneuverability. It provides a new, inovative technique in grazing ecology to explain the complicated problem of barren patch appearance in grazing grassland.关键词
秃斑/图像处理/形状指数/分形维数/特征提取Key words
barren patch/ image processing/ shape index/ fractal dimension/ feature extraction分类
农业科技引用本文复制引用
王钊齐,林慧龙..基于分形维数的放牧地秃斑特征提取方法及其应用[J].草业学报,2011,20(2):156-162,7.基金项目
国家自然科学基金重点项目(30730069)和中央高校基本科研业务费专项奖金(lzujbky-20l0-8)资助. （30730069）
