Wavelet – Feature Based-Minimum Distance vs. Rule-Based Fuzzy System for Classification of Medium Resolution Images in Heterogeneous Landscapes

Document Type : Original Article

Authors

1 Dept. of Remote Sensing and GIS, Faculty of Geography and Environmental Sciences, Hakim Sabzevari University, Sabzevar, Iran

2 Dept. of Remote Sensing and GIS, Faculty of Geography and Environmental Sciences, Hakim Sabzevari University, Sabzevar, Iran.

3 Dep. Of Geography and Urban Planning, Faculty of Geography and Environmental Sciences, Hakim Sabzevari University, Sabzevar, Iran

‎10.22052/deej.2026.258032.1129

Abstract

Classification and labeling satellite images in remote sensing (RS) as well as improvement in their accuracy have received researchers’ attentions since years ago. The present study compares two classification methods, i.e., the Rule-Based Fuzzy System and the proposed Wavelet – Feature Based-Minimum Distance (W-F-M-D) Algorithm in images with medium resolutions particularly in heterogeneous landscapes. The Advanced Land Imager (ALI) was studied in an area in southwestern of Tehran in Iran. For validation, the land cover map obtained from both methods was compared with ground truth through confusion matrix analysis, Kappa coefficient, and overall accuracy. The best results for the W-F-M-D algorithm with values of 93.55% and 0.89 were obtained for overall accuracy and Kappa coefficient. However, the results obtained from the fuzzy method for these two quantitative evaluation accuracies with values of 89.27% and 0.84 were also satisfactory. However, the simplicity and speed of the proposed W-F-M-D algorithm is another advantage over the fuzzy method. From a different point of view, in the heterogeneous urban agriculture area with a moderate spatial resolution, the accuracy obtained on the map of the urban area, as compared to the bare lands, in the W-F-M-D method in the Prod. Acc. and User. Acc. with values of 99.25% and 91.67% are evaluated as satisfactory.

Keywords

Main Subjects

Files

Share

How to cite

Statistics