Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Powiadomienia systemowe

Sesja wygasła!

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: heuristic rules

Sortuj według:

Ogranicz wyniki do:

Improved Heuristic Rules for the Price Algorithm of Global Optimization

Keil F. J., Stoyanov S. K., Yordanov S. Y.

Journal of Applied Computer Science

2007

Vol. 15, nr 1

63-82

The present study is aiming at the determination of the influence of some new heuristic rules and parameters which are proposed to improve the convergence of the Price random search procedure for global optimization. The algorithms are tested for twenty complex problems with multimodal objective functions to get reliable statistical information about the probability to find the global optimum using several heuristic procedures. The following parameters have been investigated: the type of random generator (single or multiple), the number of randomly generated initial points; the distribution of initial points and the number of cluster points. New heuristic rules are proposed for the calculation of new points in the clusters and the selection of the cluster pole. Comparison of the new algorithms is made on the basis of convergence and average number of function evaluations for the problems solved. It is proved that the modified algorithms with new heuristic rules have a better convergence and are more efficient in respect of the average number of function evaluations. The proper number of points in the initial cluster is recommended.

Pixel-based skin color detection technique

Zaqout I., Zainuddin R., Baba S.

Machine Graphics and Vision

2005

Vol. 14, No. 1

61-70

One of the simplest features used for the human face detection problem is the skin color information. A simple and relatively efficient histogram-based algorithm to segment skin pixels from a complex background is presented. The histogram-based algorithm used here is referred to as the lookup table (LUT) and is adopted to identify those intervals which may fall in the skin locus plane. For that purpose, a total of 306,401 skin samples are manually collected from RGB color images to calculate three lookup tables based on the relationship between each single pair of the three components (R, G, B). To estimate the skin locus boundary, a skin classifier box is created by integration of the proposed three heuristic rules based on how often each RGB pixel-relationship falls into its interval.