594 words - 2 pages

The amount of data generated by ultraspectral sounders is so large that considerable savings in data storage and transmission bandwidth can be achieved using data compression. Due to the large amount of data, the data compression time is of importance. Increasing programmability of commodity Graphics Processing Units (GPUs) allows their usage as General Purpose computation on Graphical Processing Units (GPGPU). GPUs offer potential for considerable increase in computation speed in applications that are data parallel. Data parallel computation on image data executes the same program on many image pixels on parallel. We have implemented a spectral image data compression method called Linear Prediction with Constant Coefficients (LP-CC) using Nvidia's CUDA parallel computing architecture. LP-CC compression method represents current the state-of-the-art in lossless compression of ultraspectral sounder data having an average compression ratio of 3.39 on publicly available NASA AIRS data. CUDA is a parallel programming architecture that is designed for data-parallel computation. CUDA does not require the programmers to explicitly manage threads. This simplifies the programming model. Our GPU implementation on Nvidia 9600 GT is experimentally compared to the native Central Processing Unit (CPU) implementation. We achieved a speed-up of 86 for an image size 135 ×90 ×2107 , when compared to a single threaded CPU version and including the data transfers between CPU and GPU. Thus, a commodity GPU can significantly decrease the computational time of a compression algorithm based on constant coefficient linear prediction.

Ultraspectral sounders generate a very large amount of data daily. Therefore, considerable savings in data storage and transmission bandwidth can be achieved using data compression. Parallel processing enables the use of computationally intensive algorithms. Previously, Message Passing Interface (MPI) and OpenMP have been used for parallel compression of hyperspectral images in [4]. A comparison of clusters and Field-Programmable Gate Arrays...

Get inspired and start your paper now!