Artificial intelligence

Scientific journal

ISSN 2710-1673

ONLINE: ISSN 2710-1681

Select your language


Особливості реалізації алгоритму AdaBoost для виявлення об’єктів на зображеннях

Murygin K.1
1 Institute of Artificial Intelligence MES of Ukraine and NAS of Ukraine

Full text (PDF)

UDC: 004.89,004.93
Publication Language: Russian
Stuc. intelekt. 2009; 14(3):573-581

Abstract: In article features of implementation of AdaBoost algorithm for solution a problem of construction of the classifiers cascade for effective detection of objects on images are considered. As elementary classifiers it is offered to use rectangular logical properties that allow to make independent a result of recognition from monotonously increasing conversions of brightness of the images which don’t change their class association. The cascade received as a result of training process 4 classifiers which allows to solve the search task of face areas in a videostream with parameters of 30 frames of the size 640×480 pixels per second.

Keywords:

References:

View full text (PDF)