研究目的
To propose a case classification method using a land cover classification case library for high-resolution remote sensing images, applying case-based reasoning to several aspects of the approach.
研究成果
The case-based reasoning model is highly suitable for representing the characteristics of land cover classification objects in high-resolution remote sensing images. The method's effectiveness was verified by extracting land cover classification information with good accuracy. The quality of information extraction is expected to improve as the case library is continually expanded and enriched.
研究不足
The selection of case objects mainly depends on manual selection, which reduces the case calculation and retrieval efficiency. The case feature selection algorithm needs further improvement to minimize the number of matching features in each type of case. The introduction of a deep learning mechanism could improve the efficiency of case updates and the accuracy of case matching.
