Abstract—Edges play a crucial role in digital image
processing. In this paper, an improved simplified novel method
is presented to improve the quality of edge detection in
grayscale images. In this method, a 3x3 binary matrix is
employed to process each and every pixel by means of vicinity.
Each pixel is considered an input and by examining binary
matrix, the edge pixel is specified and by utilizing conditions
and adaptive thresholding of the algorithm, edges are
displayed more clearly. This method not only detects edge in
images but removes noise as well. A comparative study of
different edge detection techniques like Sobel, Prewitt, Roberts,
LoG, zero-cross, Canny, fuzzy techniques and SNM with the
proposed technique is presented. The proposed technique gives
better results in terms of smoothness, sharpness, continuity,
true edges and execution time than the conventional edge
detection techniques.
Index Terms—Edge detection, digital image processing,
additive noise, image segmentation, adaptive thresholding,
morphological processing.
Tirath P. Sahu is with the Department of IT, NIT Raipur, India (e-mail:
tirasahu.it@nitrr.ac.in).
Yogendra K. Jain is with Department of CSE, SATI Vidisha, India (email:
ykjain_p@yahoo.co.in).
[PDF]
Cite:Tirath P. Sahu and Yogendra K. Jain, "Improved Simplified Novel Method for Edge Detection in Grayscale Images Using Adaptive Thresholding," Journal of Advances in Computer Networks vol. 3, no. 2, pp. 157-161, 2015.