Parallelizing an edge detection algorithm for image recognition to classify paddy and weeds leaf on Sun Fire Cluster system

  • Authors:

  • Mohd Azam Osman;Muqhtar Yassin Mohamad;Rosni Abdullah

  • Affiliations:

  • School of Computer Sciences, Universiti Sains Malaysia, Pulau Pinang, Malaysia;School of Computer Sciences, Universiti Sains Malaysia, Pulau Pinang, Malaysia;School of Computer Sciences, Universiti Sains Malaysia, Pulau Pinang, Malaysia

  • Venue:
  • SEPADS'08 Proceedings of the 7th WSEAS International Conference on Software Engineering, Parallel and Distributed Systems
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Paddy is grown widely in Malaysia. The problem of paddy growing is that weeds tend to grow with the paddy. Image recognition and classification application can be developed based on image taken to distinguish between the paddy and the weeds so that pesticide spraying and forecasting for the yield can be done. However the challenge now is that how efficient is the classification when the image is huge. In this paper, a discussion on the parallelisation of image recognition algorithm will be presented. The methodology of parallelising the Canny Edge Detection algorithm will be explained. Issues discussed includes parallelization strategy for the edge detection, followed by the implementation and the results.