In landslide analysis it is necessary to first define and extract slope units from remotely sensed data. When this is done using
hydrological process analysis, there can be complications caused by apparent sudden changes in slope gradient, resulting in the
assumption of homogeneity of regional slope stability not being met. This issue is addressed in this paper with the proposal of a
novel slope unit extraction method, which is derived from morphological image analysis (MIA) and results in the extraction of
homogeneous slope units (HSU). It is thus referred to as the MIA-HSU method. In this method, slope units are defined as small,
closed and homogeneous in 3D space. The MIA method uses logical algorithms, such as expansion and erosion, to extract slope
units. Jiangjia Gully, Yunnan Province, China was selected to test the MIA-HSU method. Results indicate that slope units
extracted by this method overcome defects of sudden changes in slope gradient and are able to identify geographical features
meaningful for further landslide analysis. The MIA-HSU method presents a more uniform slope and aspect, and is better suited to
regional landslide analysis than data obtained using the hydrological process analysis method