It is more important for spatial imformation discription of highresolution remote sensing image to improve the precision of textural features choosing. In this study, the factors to influence the textural features choosing were analyzed and the results showed that the moving window size was the main factor to affect the obtaining processes of textural features based on the Gray Level Cooccurrance Matrix method. For the highresolution remote sensing image, the most proper moving window size was determined from 9×9 to 15×15. By the calculation of OIF values, the textural features of ME，HO and CR were chosen to be best combination for the size of 3×3 and 5×5, and the combination of VA，HO and CR was considered the most properly for the moving window size from 7×7 to 17×17. Finally, if the moving window size was chosen larger than 17×17, the best combination of VA，CO and SM can be chosen. Precision assessment of different textural combinations showed that VA,HO,CR combination with optimal moving window size(from 9×9 to 17×17) could evidently improve the classification precision for highresolution remote sensing image.
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
References
[1]李金莲,刘晓玫,李恒鹏. SPOT5影像纹理特征提取与土地利用信息识别方法[J]. 遥感学报,2006,10(6):926-931.
[2]Kiema J B K. Texture analysis and data fusion in the extraction of topographic objects from satellite imagery[J]. International Journal of Remote Sensing, 2002, 23(4): 767-776.
[3]Marceau D J, Howarth P J. Evaluation of the gray level cooccurrence matrix method for land cover classification using SPOT imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 1990(28): 513-519.
[4]胡伏原,张艳宁,薛笑荣. 基于树型小波和灰度共生矩阵的SAR图像分类[J]. 系统工程与电子技术,2003,25(10):1286-1288.
[5]Sakari T, Anssi P. Performance of different spectral and textural aerial photograph features in multisource forest inventory[J]. Remote Sensing of Environment, 2005(94): 256-268.
[6]Unser M, Eden M. Multiresolution feature extraction and selection for texture sgementation[J]. IEEE Trans Pattern Anal, 1989, 2: 717-728.
[7]Kashyap R, Khoatnazd A. A model based method of rotation invariant texture classifieation[J]. IEEE Trans Pattern Analysis and Maehine Intelligenee, 1986, 8: 472-481.
[8]Pentland A. Fractal based description of natural scenes[J]. IEEE Tarns Pattern Analysis and Machine Intelligenee, 1984, 6(6): 661-674.
[9]Khotanazd A, Kashyap R. Feature selection for texture recognition based on image synthesis[J]. IEEE Transactions on Systems, 1987, 17: 1087-1095.
[10]姜青香,刘慧平,孙令彦. 纹理分析方法在TM图像信息提取中的应用[J]. 遥感信息,2003(4):72-78.
[11]卢健,彭嫚,卢昕. 遥感图像相关性及其熵计算[J]. 武汉大学学报:信息科学版,2006,6(31):476-480.
[12]吴均,赵忠明. 利用基于小波的尺度共生矩阵进行纹理分析[J]. 遥感学报,2001,5(2):28-33.
[13]孙建国,杨树文,段焕娥,等. 基于光谱和纹理特征的山区高分辨率遥感影像分类[J]. 测绘科学,2009,34 (6):92-93.
[14]罗火钱,齐宇,谢爱英. 不同融合算法对ETM+遥感影像分类精度的影响分析[J]. 吉林师范大学学报:自然科学版,2009,4:48-54.
[15]黄昕,张良培,李平湘. 基于小波的高分辨率遥感影像纹理分类方法研究[J]. 武汉大学学报:自然科学版,2006,31(1):66-69.
PDF(1533530 KB)
