Комп'ютерні науки та інженерія (CSE-2011 ). – 2011 р.

Permanent URI for this collectionhttps://ena.lpnu.ua/handle/ntb/22503

Browse

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Параметри оптимізації у медицині з підтримкою алгоритмів розпізнання зображення
    (Видавництво Львівської політехніки, 2011) Бржескі, Адам
    In this paper, a procedure of automatic set up of image recognition algorithms' parameters is proposed, for the purpose of reducing the time needed for algorithms' development. The procedure is presented on two medicine supporting algorithms, performing bleeding detection in endoscopic images. Since the algorithms contain multiple parameters which must be specified, empirical testing is usually required to optimise the algorithm's efficiency, which is time-consuming and therefore costly. To approach this problem, the parameters were transformed into a solution space, which was later searched using various local and global optimization methods. After the optimization was performed, the achieved efficiency rates were compared to the empirically established values, showing similar effects, and therefore successfulness of the automatic parameters optimization.
  • Thumbnail Image
    Item
    Розпаралелювання алгоритмів відеопотоку в платформі Каскада
    (Видавництво Львівської політехніки, 2011) Бржескі, Адам
    The purpose of this work is to present different techniques of video stream algorithms parallelization provided by the Kaskada platform - a novel system working in a supercomputer environment designated for multimedia streams processing. Considered parallelization methods include frame-level concurrency, multithreading and pipeline processing. Execution performance was measured on four time-consuming image recognition algorithms, designed to support medical endoscopic examinations. The achieved results confirm high capabilities of Kaskada platform for executing parallel algorithms. Frame-level parallelization was proved to be a great solution for offline processing, while pipeline processing combined with multithreading provided high performance for online, real-time analysis.