Wyniki wyszukiwania - BazTech

Ograniczanie wyników

2 2019

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: cross entropy

Sortuj według:

Ogranicz wyniki do:

Application of convolutional neuron network for image processing and interpretation

Pałczyński Krzysztof

Zeszyty Naukowe. Telekomunikacja i Elektronika / Uniwersytet Technologiczno-Przyrodniczy w Bydgoszczy

2019

nr 23

5--12

This article describes the application of Convolutional Neural Network in image processing and describes how it works. There are presented: network layers, types of activation functions, example of the AlexNet network architecture, the use of the loss function and the cross entropy method to calculate the loss during tests, L2 and Dropout methods used for weights regularization and optimization of the loss function using Stochastic Gradient Drop.

Artykuł ten opisuje zastosowanie Konwolucyjnych Sieci Neuronowych w przetwarzaniu obrazów. W celu lepszego zrozumienia tematu opisano sposób działania sieci. Przedstawiono sieci wielowarstwowe, rodzaje funkcji aktywacji, przykład architektury sieci AlexNet. W artykule skupiono się na opisaniu wykorzystania funkcji straty oraz metody entropii krzyżowej do obliczenia straty w czasie testów. Opisano również sposoby normalizacji wag L2 i Dropout oraz optymalizację funkcji straty za pomocą Stochastycznego Spadku Gradientu.

A hybrid method for blood vessel segmentation in images

Díaz Primitivo, Rodríguez Alma, Cuevas Erik, Valdivia Arturo, Chavolla Edgar, Pérez-Cisneros Marco, Zaldívar Daniel

Biocybernetics and Biomedical Engineering

2019

Vol. 39, no. 3

814--824

In the last years, image processing has been an important tool for health care. The analysis of retinal vessel images has become crucial to achieving a better diagnosis and treatment for several cardiovascular and ophthalmological deceases. Therefore, an automatic and accurate procedure for retinal vessel and optic disc segmentation is essential for illness detection. This task is extremely hard and time-consuming, often requiring the assistance of human experts with a high degree of professional skills. Several retinal vessel segmentation methods have been developed with satisfactory results. Nevertheless, most of such techniques present a poor performance mainly due to the complex structure of vessels in retinal images. In this paper, an accurate methodology for retinal vessel and optic disc segmentation is presented. The proposed scheme combines two different techniques: the Lateral Inhibition (LI) and the Differential Evolution (DE). The LI scheme produces a new image with enhanced contrast between the background and retinal vessels. Then, the DE algorithm is used to obtain the appropriate threshold values through the minimization of the cross-entropy function from the enhanced image. To evaluate the performance of the proposed approach, several experiments over images extracted from STARE, DRIVE, and DRISHTI-GS databases have been conducted. Simulation results demonstrate a high performance of the proposed scheme in comparison with similar methods reported in the literature.