Performance Evaluation of Different Images Using Edge Detection Algorithms

Authors

  • Chhetra Bahadur Chhetri Lecturer of NCCS, Paknajol, Kathmandu, Nepal
  • Manish Pandey Student of BCA 6th semester, NCCS, Nepal

DOI:

https://doi.org/10.3126/nccsrj.v2i1.60053

Keywords:

Canny Edge detection, Log Edge detection, cycle/byte, image processing, empirical performance

Abstract

To determine which edge detection method performs best and worst on different image types, numerous edge detection algorithms are examined. For the performance analysis, some sample photos from the web and some from Java are used as sources. The entropy and signal noise ratio are used to gauge how well the edged image performs. In image processing, conducting a thorough investigation of various edge detection techniques is highly worthwhile two widely used edge detection algorithms Log, and Canny—are taken into consideration in this analysis. Here in this paper, the analysis is focused on the performance of different edge detector algorithms. All candidate algorithms of edge detection are implemented in JAVA. The result of empirical performance shows that two variants namely canny perform better results for the edge detection algorithm. The result shows that when considering only the performance aspect. Cycle/byte is calculated for comparing different variants. Cycle/byte is decreased when the canny edge detector is examined. The canny edge detection algorithm shows a better performance than LoG. LoG has more than 3 times higher cycle/byte than Canny Edge detection.

Downloads

Download data is not yet available.
Abstract
46
PDF
43

Downloads

Published

2023-11-27

Issue

Section

Articles