Evaluation of multi-temporal/sensor data fusion for urban change analysis

Natural Resource Management, Planning and Monitoring programs depend on accurate information about the land cover/use for varying sized regions. Satellite sensor images with moderate to various resolutions have facilitated scientific research activities at landscape and regional scales. In this study, a part of Istanbul metropolitan area that faced a great land cover/use change was investigated. For that purpose, 1992 dated SPOT 4 panchromatic image was fused with 2017 dated Landsat 8 OLI multispectral data. As an inverse approach 2017 dated Landsat 8 OLI Panchromatic image was fused with 1984 dated Landsat 5 TM multispectral data. Image fusion processes was performed using Gram Schmidt spectral sharpening algorithm which integrates all bands of the multispectral image into fusion. After fusion operations, two image datasets were classified in order to determine land cover/use changes. Results of the study showed that proposed methodology provided efficient and rapid determination of the changes due to their distinct spectral characteristics from their surroundings in fused images. These methods add significant advantages for land use/cover classification.