Comparison Of Various Feature Extraction Methods In Digit Recognition


YILDIZ E., SEVİM Y.

26th IEEE Signal Processing and Communications Applications Conference (SIU), İzmir, Turkey, 2 - 05 May 2018, (Full Text) identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/siu.2018.8404667
  • City: İzmir
  • Country: Turkey
  • Karadeniz Technical University Affiliated: Yes

Abstract

Digit recognition can be difficult due to different writing styles of people. Therefore, there are a lot of feature extraction methods. In this paper, commonly used feature extraction methods in digit recognition and their performance are investigated. These methods are Gradient Features, Diagonal Features, Zoninig Features, Projection Histograms Features and Chaincode Features.