An Evaluation of the Use of Integrated Spectral and Textural Features to Identify Agricultural Land Cover Types in Pangalengan, West Java, Indonesia

JARQ : Japan Agricultural Research Quarterly
ISSN 00213551
書誌レコードID(総合目録DB) AA0068709X
本文フルテキスト
38-02-10.pdf1.42 MB

The objective of the study is to evaluate the use of integration of spectral and textural features derived from IKONOS imagery to identify agricultural land cover types in a mountainous case study area in Pangalengan, West Java, Indonesia. The study includes image preprocessing, development of an image quantization method, calculation of textural measures, development of data sets and an accuracy assessment. Image preprocessing focuses on image registration and topographic normalization. Topographic normalization is conducted to minimize the effect of illumination differences on surface reflectance. In this study, two image quantization methods, i.e. image segmentation and averaging filtered were developed. The image segmentation method classifies the image into several segmentations based on a determination of the total number of pixels per class, while the averaging filtered method classifies the image based on the average of the digital number values within a window size. Four textural measures, inverse difference moment, contrast, entropy and energy, were calculated based on the gray level co-occurrence matrix (GLCM). The results indicate that a combination of spectral and textural aspects significantly improves the classification accuracy compared with classification with pure spectral features only. Image segmentation and averaging filtered methods can reveal more effectively spatial forms of agricultural land cover types than using a 256 gray-level scale. The overall accuracy increased 11.33% when using the integration of spectral and multiple textural features of inverse difference moment (5×5) and energy (9×9).

刊行年月日
作成者 WIKANTIKA Ketut UCHIDA Satoshi YAMAMOTO Yukiyo
著者キーワード

land cover classification

image quantization

IKONOS

公開者 Japan International Research Center for Agricultural Sciences
オンライン掲載日
国立情報学研究所メタデータ主題語彙集(資源タイプ) Journal Article
38
2
開始ページ 137
終了ページ 148
DOI 10.6090/jarq.38.137
権利 Japan International Research Center for Agricultural Sciences
言語 eng

関連する刊行物