Image Segmentation of Pseudo-foreign Fibers in Cotton on the Basis of Improved Genetic Algorithm - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2012

Image Segmentation of Pseudo-foreign Fibers in Cotton on the Basis of Improved Genetic Algorithm

Lulu Ge
  • Fonction : Auteur
  • PersonId : 988286
Daoliang Li
  • Fonction : Auteur
  • PersonId : 988287
Liu Yang
  • Fonction : Auteur
  • PersonId : 972250
Wenzhu Yang
  • Fonction : Auteur
  • PersonId : 972288

Résumé

In the foreign fibers cleaning process, pseudo-foreign fibers are often mistaken for foreign fibers, this result not only seriously affects the detecting precision of foreign fibers cleaning machine, but also doubles the time of cleaning up lint. As for false identification problem of pseudo-foreign fibers in cotton, this paper proposes a new approach for fast segmentation of pseudo-foreign fibers in cotton on the basis of improved genetic algorithm. This improved genetic algorithm reduced the searching range for calculating optimal threshold from 0~255 to 100~220. The calculating speed in this stage was improved more than twice in average. The fitness amendments formula is also proposed to improve genetic algorithm disadvantage, at the same time, this solved issues of "premature", and converging to global optimal solution difficultly in the traditional algorithm. The results show that the algorithm has high speed, accuracy, anti-interference and so on.
Fichier principal
Vignette du fichier
978-3-642-27278-3_55_Chapter.pdf (4 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01361027 , version 1 (06-09-2016)

Licence

Paternité

Identifiants

Citer

Lulu Ge, Daoliang Li, Liu Yang, Wenzhu Yang. Image Segmentation of Pseudo-foreign Fibers in Cotton on the Basis of Improved Genetic Algorithm. 5th Computer and Computing Technologies in Agriculture (CCTA), Oct 2011, Beijing, China. pp.538-548, ⟨10.1007/978-3-642-27278-3_55⟩. ⟨hal-01361027⟩
68 Consultations
80 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More