Landscape images distance using kullback leibler divergence Documento de conferencia uri icon

Abstracto

  • We present in this document a new proposal to measure distance for landscape images taking into consideration color, texture and, orientation. The hue, entropy and gradient field's magnitude and orientation are used to define features within the images. Distance measurement between images of different sizes were performed using histograms and the Kullback Leibler divergence. The results show that the proposed way to measure the distance performance can identify differences between the landscapes analyzed: forest, city and, dessert.

fecha de publicación

  • 2018

Palabras clave

  • Entropy
  • Forestry
  • Image color analysis
  • Urban areas
  • hsl color
  • Computational modeling
  • Histograms
  • Kullback Leibler
  • Mathematical model
  • image distance