Design and Storage Optimization of GPU-based Parallel Program of Image Registration for Remote Sensing

Image registration is a crucial step of many remote sensing related applications. As the scale of data and complexity of algorithm keep growing, image registration faces great challenges of its processing speed. In recent years, the computing capacity of GPU improves greatly. Taking the benefits of using GPU to solve general propose problem, we research on GPU-based remote sensing image registration algorithm. A mutual information based wavelet registration algorithm is proposed on the GPU parallel programming model, and storage optimization strategy is applied on the registration process. Using CUDA language, we tested our proposed methods with nVIDIA Tesla M2050 GPU. The experiment results prove that the parallel programming model and storage optimization strategy can well adapt to the field of remote sensing image registration, with a speedup of 19.9x. Our research also shows that the GPU-based general propose computing has a bright future in the field of remote sensing image processing.