A Machine Learning-Based Approach for Land Cover Change Detection Using Remote Sensing and Radiometric Measurements

An approach combining the Hotelling $T^{2}$ control method with a weighted random forest classifier is proposed and used in the context of detecting land cover changes via remote sensing and radiometric measurements. Hotelling $T^{2}$ procedure is introduced to identify features corresponding to changed areas. Nevertheless, $T^{2}$ scheme is not able to separate real from false changes. To tackle this limitation, the weighted random forest algorithm, which is an efficient classification technique for imbalanced problems, has been successfully applied to the features of the detected pixels to recognize the type of change. The feasibility of the proposed procedure is verified using SZTAKI AirChange benchmark data. Results proclaim that the proposed detection scheme succeeds to effectively identify land cover changes. Also, the comparisons with other methods (i.e., neural network, random forest, support vector machine, and $k$ -nearest neighbors) highlight the superiority of the proposed method.