Desarrollo de un algoritmo que permita paralelizar imágenes 3D sobre tecnología Nvidia y OpenCL

The present project is developed around the problem about the processes that are executed serially in the CPU and the excessive time it takes to show results of the user mainly in the image processing. The offer of the solution to improve the execution time is the parallelization of processes and sp...

Descripción completa

Autor Principal: Párraga Párraga, Jéssica Virginia
Otros Autores: Casa López, Luis Fernando
Formato: bachelorThesis
Idioma: spa
Publicado: 2017
Materias:
Acceso en línea: http://dspace.ups.edu.ec/handle/123456789/14117
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Sumario: The present project is developed around the problem about the processes that are executed serially in the CPU and the excessive time it takes to show results of the user mainly in the image processing. The offer of the solution to improve the execution time is the parallelization of processes and spoils them in the GPU, for which NVIDIA technology is used and OpenCL programming. Regarding to image processing, the Retinex algorithm is applied, from which it can be emphasized that it is based on the human vision to improve the colors and scenarios of low luminosity. DataSet of computerized tomographies was used for the tests performed, which allow to display a 3D environment with the information of the frames that it contains.