MODIFIED METHOD FOR DETECTING SMALL STRUCTURES ON NOISY IMAGES
A. V. Likhachev
Institute of Automation and Electrometry, Siberian Branch, Russian Academy of Sciences, Novosibirsk, Russia
Keywords: классификатор элементов изображения, апостериорные вероятности, image element classifier, posterior probabilities
Abstract
A classifier for operation with 1 pixel image fragments is proposed. Separation of elements into an object and background is based on posterior probabilities, which are calculated using a histogram. A computational experiment has shown that the developed algorithm provides a more accurate classification than the separation according to the brightness threshold determined from the condition of the minimum weighted sum of errors of the first and second kind.
|
