Распознавание объекта в сцене на основе модели непроизвольного внимания с использованием OpenCL

Detalles Bibliográficos
Parent link:Технологии Microsoft в теории и практике программирования.— 2014.— [С. 97-99]
Autor Principal: Потапьев И. А.
Autor Corporativo: Национальный исследовательский Томский политехнический университет (ТПУ) Институт кибернетики (ИК) Кафедра оптимизации систем управления (ОСУ)
Outros autores: Аксёнов С. В. Сергей Владимирович (727)
Summary:Заглавие с титульного листа.
In this paper the algorithm of object tracking based on bottom-up attention is presented. First, the one-dimensional images were extracted from original image such as color and intensity maps. Then they combined into saliency map, which maximums corresponds to the most salient locations. Finally, the adaptive threshold is used to determine the location of objects on image. To accelerate the algorithm, the OpenCL version of the algorithm has been implemented. Experimental results demonstrate effectiveness of the algorithm for single and multiple object tracking. Implementation on OpenCL improves performance of the system.
Idioma:ruso
Publicado: 2014
Series:Математическое моделирование и технологии высокопроизводительных вычислений
Subjects:
Acceso en liña:http://www.lib.tpu.ru/fulltext/c/2014/C28/037.pdf
Formato: Electrónico Capítulo de libro
KOHA link:https://koha.lib.tpu.ru/cgi-bin/koha/opac-detail.pl?biblionumber=609520
Descripción
Summary:Заглавие с титульного листа.
In this paper the algorithm of object tracking based on bottom-up attention is presented. First, the one-dimensional images were extracted from original image such as color and intensity maps. Then they combined into saliency map, which maximums corresponds to the most salient locations. Finally, the adaptive threshold is used to determine the location of objects on image. To accelerate the algorithm, the OpenCL version of the algorithm has been implemented. Experimental results demonstrate effectiveness of the algorithm for single and multiple object tracking. Implementation on OpenCL improves performance of the system.