A Modified Graph Based Approach for Leaf Segmentation with GPGPU Support

Donmez E., Zadeh P. V.

23nd Signal Processing and Communications Applications Conference (SIU), Malatya, Turkey, 16 - 19 May 2015, pp.1797-1800 identifier

  • Publication Type: Conference Paper / Full Text
  • City: Malatya
  • Country: Turkey
  • Page Numbers: pp.1797-1800
  • Istanbul Technical University Affiliated: Yes


Image segmentation is a hard problem to solve object detection from an image before classify process. There are several techniques which are implemented with respect to what will be segmented from image. The aim of this script is extracting any leaf from its background by using segmentation techniques as possible as sufficient for identification. There is a point which has to be emphasized; the backgrounds of our leaf data set are not white, all leaf data set backgrounds are complex nature component; such as a forest, a river, sky, and a group of mountains or etc. In this study, we have implemented several techniques to segment leaf from its background throughout our analysis, and eventually we have proposed a graph based robust segmentation process for leaf segmentation.