Detection of curved lines and estimation of their parameters on images
Autor: | Pavel V. Babayan, Nikita Yu. Shubin |
---|---|
Rok vydání: | 2017 |
Předmět: |
Radon transform
Computer science business.industry 02 engineering and technology Curvature 01 natural sciences Image (mathematics) 010309 optics 020210 optoelectronics & photonics 0103 physical sciences 0202 electrical engineering electronic engineering information engineering Computer vision Artificial intelligence business General algorithm |
Zdroj: | MECO |
DOI: | 10.1109/meco.2017.7977196 |
Popis: | An approach to effective detection and parameters estimation of long edges in the image has been proposed. The main idea of this work is the detection of long edges by parts, which curvature can be neglected. In accordance with the general algorithm the image is divided into blocks which are searched for straight lines. Then in-block and interblock filtering to reduce the number of false detections and combining the individual segments into long curved lines are used. The interblock filtering based on multi-agent approach. |
Databáze: | OpenAIRE |
Externí odkaz: |