Automatic road extraction based on neuro-fuzzy algorithm

  • Authors:

  • Asef Zare;Mostafa Okauti

  • Affiliations:

  • Electrical Engineering, Islamic Azad University, Gonabad, Iran;Electrical Engineering, Islamic Azad University, Gonabad, Iran

  • Venue:
  • ROCOM'10 Proceedings of the 10th WSEAS international conference on Robotics, control and manufacturing technology
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, presents a automatic method based on neuro-fuzzy algorithm for extraction of road in satellite image. This method is compared to a fuzzy algorithm. The method consist of three steps; feature extraction (preprocessing), neuro-fuzzy system modeling and post processing. These methods are applied to extract the roads from a satellite image. The input parameters for this system are mean and standard deviation of the 5 × 5 windows. Initial parameters are selected and compared to a selected database made of 20 satellite images. In this paper, for better detection and extraction of roads, a fuzzy network is trained through a neural network. The results show that, the neurofuzzy algorithm gives much better performance than the fuzzy algorithm. The main features of the proposed model are that it is automatic method and have the minimum error rate and maximum accuracy.