Development of Key Image Processing Methods and Environmental Parameter Models for Formosat-5 Data and Applications( I )

Formosat-2 was launched in 2004, and was the first Taiwan’s remote sensing satellite. The onboard RSI equipped with 8 meter and 2 meter spatial resolution for multispectral and panchromatic images respectively. Before it was dimissioned in 2016, it accumulated a huge amount of images for various applications including land cover, land use, natural resources, environmental monitoring and disaster management. The following Formosat-5 launched in 2017, was the first optical remote sensing satellite developed by ourselves. It has the same spatial resolution of panchromatic image with Formosat-2, but better resolution of 4 meter in multispectral image. However, because of the focus problem, the image quality from Formosat-5 was not very good. NSPO spends efforts on image processing technics to resolve the blur issue focus on images themselves. In this proposal, we will take a different approach by focusing on the applications. Center for Space and Remote Sensing Research has worked on remote sensing technology and geoinformatics for more than two decades with experiences and strong team. Our proposal is based on applications, to improve the quality of radiometric, spectral and spatial information, and the coregistration between bands. Image processing techniques will also be designed for image fusion and mosaic, color balance, geographic calibration and cloud/haze removal. Finally, the historical images from Formosat-2 will be include to form multi-temporal image set for environmental parameter retrieval and change detection.