2 sonuçlar
Arama Sonuçları
Listeleniyor 1 - 2 / 2
Yayın Robust localization and identification of African clawed frogs in digital images(Elsevier Science BV, 2014-09) Tek, Faik Boray; Cannavo, Flavio; Nunnari, Giuseppe; Kale, İzzetWe study the automatic localization and identification of African clawed frogs (Xenopus laevis sp.) in digital images taken in a laboratory environment. We propose a novel and stable frog body localization and skin pattern window extraction algorithm. We show that it compensates scale and rotation changes very well. Moreover, it is able to localize and extract highly overlapping regions (pattern windows) even in the cases of intense affine transformations, blurring, Gaussian noise, and intensity transformations. The frog skin pattern (i.e. texture) provides a unique feature for the identification of individual frogs. We investigate the suitability of five different feature descriptors (Gabor filters, area granulometry, HoG,(1) dense SIFT,(2) and raw pixel values) to represent frog skin patterns. We compare the robustness of the features based on their identification performance using a nearest neighbor classifier. Our experiments show that among five features that we tested, the best performing feature against rotation, scale, and blurring modifications was the raw pixel feature, whereas the SIFT feature was the best performing one against affine and intensity modifications.Yayın Malaria parasite detection with deep transfer learning(IEEE, 2018-12-06) Var, Esra; Tek, Faik BorayThis study aims to automatically detect malaria parasites (Plasmodium sp) on images taken from Giemsa stained blood smears. Deep learning methods provide limited performance when sample size is low. In transfer learning, visual features are learned from large general data sets, and problem-specific classification problem can be solved successfully in restricted problem specific data sets. In this study, we apply transfer learning method to detect and classify malaria parasites. We use a popular pre-trained CNN model VGG19. We trained the model for 20 epoch on 1428 P Vivax, 1425 P Ovule, 1446 E Falciparum, 1450 P Malariae and 1440 non-parasite samples. The transfer learning model achieves %80, %83, %86, %75 precision and 83%, 86%, 86%, 79% f-measure on 19 test images.
