Performances Study of Tomographic Reconstruction Implemented on Nvidia Graphic Processor Systems
نویسندگان
چکیده
Procesul de reconstrucţie tomografică necesită un cost computaţional foarte mare datorită procedurii de “proiecţie înapoi” (backprojection). În acest articol am implementat un astfel de algoritm de reconstrucţie pentru o geometrie paralelă 2D a razelor X. Implementarea a fost realizată în 4 variante: utilizând o singură unitate de calcul, 4 unităţi de calcul (Intel Q6600 la 2.4 GHz şi 8GB memorie ram) împreună cu tehnologia OpenMp, o placă video ce este compatibilă cu tehnologia CUDA şi utilizează memoria globală a acesteia şi ultima variantă aceeaşi placă video însă de data aceasta utilizând memoria texturilor. Atât implementările pe CPU, cât şi cele pe placa grafică folosesc datele în simplă precizie. Cel mai bun timp de calcul a fost obţinut în cazul utilizării plăcii video cu memoria texturilor.
منابع مشابه
Fast Cellular Automata Implementation on Graphic Processor Unit (GPU) for Salt and Pepper Noise Removal
Noise removal operation is commonly applied as pre-processing step before subsequent image processing tasks due to the occurrence of noise during acquisition or transmission process. A common problem in imaging systems by using CMOS or CCD sensors is appearance of the salt and pepper noise. This paper presents Cellular Automata (CA) framework for noise removal of distorted image by the salt an...
متن کاملFast tomographic reconstruction on multicore computers
SUMMARY Tomo3D implements a multithreaded vectorized approach to tomographic reconstruction that takes full advantage of the computer power in modern multicore computers. Full resolution tomograms are generated at high speed on standard computers with no special system requirements. Tomo3D has the most common reconstruction methods implemented, namely weighted Back-projection (WBP) and simultan...
متن کاملImplementation of the r.cuda.los module in the open source GRASS GIS by using parallel computation on the NVIDIA CUDA graphic cards
Parallel computing is in expanding phase in GIS applications. A very attractive solution for parallel computing are the NVIDIA graphic cards, with a parallel computing platform and the CUDA (Compute Unified Device Architecture) programming model. The basis for this paper is the r.los module used to calculate optical visibility (LOS Line of Sight), which is already implemented in the GRASS GIS e...
متن کاملAccelerating frequency-domain diffuse optical tomographic image reconstruction using graphics processing units.
Diffuse optical tomographic image reconstruction uses advanced numerical models that are computationally costly to be implemented in the real time. The graphics processing units (GPUs) offer desktop massive parallelization that can accelerate these computations. An open-source GPU-accelerated linear algebra library package is used to compute the most intensive matrix-matrix calculations and mat...
متن کاملDetermining the Proper compression Algorithm for Biomedical Signals and Design of an Optimum Graphic System to Display Them (TECHNICAL NOTES)
In this paper the need for employing a data reduction algorithm in using digital graphic systems to display biomedical signals is firstly addressed and then, some such algorithms are compared from different points of view (such as complexity, real time feasibility, etc.). Subsequently, it is concluded that Turning Point algorithm can be a suitable one for real time implementation on a microproc...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2011