Multi temporal Land use classification using synthetic aperture radar images

Land use classification of urban area is one of the most important applications of remote sensing images. Urban management is generally in demand to account the dynamics of land uses. In this subject, multi-temporal classification accounts the timely (seasonal or annual) variation of cities, which is very useful to understand the impact of land use dynamic on urban’s morphology. Several land use classification techniques based on optical data are so far available. However, it turns out that half of the world is in nighttime, and half of the world, on average, is cloudy. When combining these two together, about 75 percent of Earth, at any given time, is going to be cloudy, nighttime, or it's going to be both. This issue urges requirements to radar remote sensing-based classification, which allows us to monitor our cities in all-weather and in all-light conditions and to do so reliably and transparently. Thus, this study will perform a multi-temporal land use classification using RadaraSAT-2 images. The selected test site for the classification is Enschede.