Welcome to visit《 Journal of Air Force Engineering University 》Official website!

Consultation hotline:029-84786242 RSS EMAIL-ALERT
A Feature Extraction Inspired by Primary Visual Cortex Mechanism
DOI:
CSTR:
Author:
Affiliation:

Clc Number:

TP391.4

Fund Project:

  • Article
  • |
  • Figures
  • |
  • Metrics
  • |
  • Reference
  • |
  • Related
  • |
  • Cited by
  • |
  • Materials
  • |
  • Comments
    Abstract:

    In view of the limitation by the traditional methods in texture feature extraction and local motion perception, and referred to the powerful information processing mechanism of human visual system, a primary visual cortex(V1)vision percention model based on 3D spatiotemporal motion energy model is proposed. Firstly, simple cell receptive field is simulated by using 3D Gabor filter, then by half wave rectification operation, and further simple cell model is obtained. Finally, the simple cell receptive field is integrated by energy model, so that complex cell model is simulated. The simulation results show that the model possess good performance at extracting static texture feature of images and the local motion feature of simple video sequences. Compared with the traditional model, the model combines computer vision well with biovision, having stronger abilities at feature representation and extraction, and showing more powerful information processing mechanism of V1 cells. And this also provides new ideas for braininspired visual computing.

    Reference
    Related
    Cited by
Get Citation
Share
Article Metrics
  • Abstract:
  • PDF:
  • HTML:
  • Cited by:
History
  • Received:
  • Revised:
  • Adopted:
  • Online: December 28,2018
  • Published:
Article QR Code