Email this article Analyzing the Efficiency of Segment Boundary Detection Using Neural Networks
- Authors: Kugaevskikh A.V.1,2,3, Sogreshilin A.A.2
-
Affiliations:
- Novosibirsk State Technical University
- Novosibirsk State University
- Institute of Automation and Electrometry, Siberian Branch
- Issue: Vol 55, No 4 (2019)
- Pages: 414-422
- Section: Analysis and Synthesis of Signals and Images
- URL: https://journals.rcsi.science/8756-6990/article/view/212827
- DOI: https://doi.org/10.3103/S8756699019040137
- ID: 212827
Cite item
Open Access
Access granted
Subscription Access
Abstract
—This paper describes the architecture of a neural network for edge detection. Different filters for first-layer neurons are compared. Neural network learning based on a cosine measure algorithm shows much worse results than an error backpropagation algorithm. Optimal parameters for the first-layer neuron operation are given. The proposed architecture fulfills the stated tasks on edge selection.
About the authors
A. V. Kugaevskikh
Novosibirsk State Technical University; Novosibirsk State University; Institute of Automation and Electrometry, Siberian Branch
Author for correspondence.
Email: a-kugaevskikh@yandex.ru
Russian Federation, pr. Karla Marksa 20, Novosibirsk, 630073; ul. Pirogova 1, Novosibirsk, 630090; pr. Akademika Koptyuga 1, Novosibirsk, 630090
A. A. Sogreshilin
Novosibirsk State University
Email: a-kugaevskikh@yandex.ru
Russian Federation, ul. Pirogova 1, Novosibirsk, 630090
Supplementary files
