计算机应用与软件2016，Vol.33Issue(9)：154-158,5.DOI:10.3969/j.issn.1000-386x.2016.09.037
基于卫星图像的古民居识别算法
RECOGNITION ALGORITHM OF ANCIENT DWELLINGS BASED ON SATELLITE IMAGES
摘要
Abstract
We study the determination of village images from satellite maps according to the scope of longitude and latitude.The dwelling objects are extracted by using rectangular algorithm based on greyscale feature and shape feature of dwellings.For those misrecognised dwellings objects,it is needed to further extract internal texture features of the object for discrimination.We use statistics value of grey-level co-occurrence matrix (GLCM)to represent the feature of every dwelling object,and form the eigenvector of dwellings image.GMLVQ classification algorithm is used to classify the extracted dwellings objects and thus the exclusion of non-dwellings can be achieved. Experiments show that the extraction rate of dwellings is up to 76.1% when the appropriate parameters are selected,such as the length and width of the rectangle,and the related threshold.Using the eigenvector classification algorithm generated based on the elastic mesh and GLCM,the correct recognition rate of dwellings is higher than 88.9%.关键词
卫星地图/矩形法/弹性网格/灰度共生矩阵/民居识别Key words
Satellite image/Rectangular algorithm/Elastic mesh/GLCM/Dwellings recognition分类
信息技术与安全科学引用本文复制引用
杨帆,沈来信,王雪飞..基于卫星图像的古民居识别算法[J].计算机应用与软件,2016,33(9):154-158,5.基金项目
文化部科技创新项目（WHBKJCXXM20142554）；文物局文化保护和技术研究课题（2013-YB-SQ-175）；省级大学生创新创业训练计划项目（AH2014103753135，AH2014103753136）。 （）
