An IHS-based change detection approach for assessment of urban expansion impact on arable land loss in China

China is experiencing a drastic loss of arable land as a result of the continuing process of rapid urbanization. This loss could have an adverse impact not only on China's sustainable development but also on the world food market. To provide a robust and economical tool for evaluating the impact of this process, we developed an IHS (Intensity, Hue, Saturation)-based approach for detecting the change from arable land to urban land by using Landsat TM and SPOT Pan images in different years. This approach takes full advantage of the colour distortion related to the changes in objects after IHS and reverse IHS transformation. Based on case studies in Guangzhou city and Nanyang city, choosing Landsat TM bands 7, 4, and 3 as the input RGB component and employing an exponential contrast stretch and a high-pass filter in SPOT Pan pre-processing led to the greatest change in detection accuracy.