Using multi layer perceptron network to classify road cracks

  • Authors:

  • Heydar Toossian Shandiz;Hosein Ghasemzadeh Tehrani;Hadi Hadizadeh

  • Affiliations:

  • Shahrood University of Technology, Electrical Engineering Faculty, Civil Engineering Faculty, Shahrood, Iran;Shahrood University of Technology, Electrical Engineering Faculty, Civil Engineering Faculty, Shahrood, Iran;Shahrood University of Technology, Electrical Engineering Faculty, Civil Engineering Faculty, Shahrood, Iran

  • Venue:
  • NN'05 Proceedings of the 6th WSEAS international conference on Neural networks
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper a method for classifying cracks in asphalt road by using multi layer Perceptron (MLP) neural network is proposed. Training data are road images which are taken from road surface in 30 degree. The RGB images are first changed to gray scale and then binary images are produced by using proper threshold gray level. The structure of network has three layers as input, hidden and output. The network is trained to perform tasks such as pattern recognition. The training rule is categorized as back propagation method and learning method is supervised. After training the network, it classifies each image to longitudinal, transverse, blocks, Alligator and others.