Natural land cover patterns continuously undergo changes, impacted by various natural as well as human-managed factors. The remotely sensed data are commonly utilized to detect land cover change, which is important to understanding long-term landscape dynamics. Generally, a methodology for global change is composed of mapping, quantifying, and monitoring changes in the physical characteristics of land cover. The selected processing and analysis techniques affect the quality of the obtained information. In this research, a change detection/feature extraction system is proposed based on remotely sensed data: preprocessing, change detection and segmentation, resulting in the mapping of the change-detected areas. Here, appropriate methods are studied for each step and in particular, in the segmentation process, a multiresolution framework to reduce computational complexity is investigated for multitemporal images of large size.