Unsupervised monitoring of urban land use and land cover change detection in multitemporal images

Remote sensing is the most reliable and effective way to monitor the land use and land cover changes (LULC). In the present day world, this LULC mapping is of great significance in scientific, scholarly research, planning and management. Based on remotely sensed data acquired on two different periods it could be quite possible to detect the changes. This paper proposes a novel unsupervised monitoring approach based on Undecimated Discrete Wavelet Transforms (UDWT). In urban areas, detection of land space destruction is highly essential. In this paper, a novel UDWT transform domain approach to construct change detection binary map is performed initially on two co- registered images. Change Vector Analysis (CVA) is applied next on the two temporal images. As a result, change or no change map is produced. Finally forecasting of the land use trend changes using markov chain is done.