Bonfring International Journal of Advances in Image Processing, Vol. 4, No. 1, December 2014 1
Abstract--- Detecting the edges of objects within images is
a critical task for quality image processing. This paper
proposes an edge detection operator based on the
combination of fuzzy gradient morphology and Sobel
operator. When we use traditional detection operators to
detect the edge of an object in an image, we get lots of noisy
Points. In this paper, we demonstrate a technique in which we
preprocess an image with Sobel operator and then apply
gradient morphology. This method effectively removes the
noise and gives good detail image edge detection. We evaluate
the method quantitatively and compare it to classical
morphological method. Our fuzzy based edge segmentation
method performs better than the classical edge detectors.
Since the proposed methods are based on fuzzy morphological
operations, these are efficient and enhanced.
Keywords--- Fuzzy, Mathematical Morphology,
Segmentation, Sobel Operator
I. INTRODUCTION
ODAY edge detection is a very important area in the field
of image processing and computer vision. It also plays an
important role for image segmentation and object recognition.
Edge detection is the process to detect the important features
of image. Here, features mean the properties of image like
discontinuities in physical and geometric characteristics of an
image or abrupt variation in its intensity, Wang et al.,[1]. The
quality of edges is affected by the presence of objects in
similar illumination, noise and density of edges,Nadarnejat
[14]. The variation in characteristics can lead to the variation
in gray level of the image. Edge detection is therefore
considered as an important step for facilitating higher level
image analysis and processing, Mittal and Batra [2] and
Gonzalez et al., [3].
Conventionally, edge was detected using gradient
algorithms like Sobel , Prewitt andLaplacian of Gaussian
operator,Huertas and Medioni [4], all of which belong to the
high pass filtering methods. Another important gradient based
edge detection method is Canny algorithm which solves an
optimization problem to detect the edges, Canny [5]. The
tradeoff between detection and location of edge pixels make a
problem in accuracy,Wang et al., [1]. By changing threshold
values, edge detection rate increases, but the accuracy of edge
Dillip Ranjan Nayak, Asst. Professor, Computer Science & Engineering,
Govt. College of Engineering, Kalahandi, Bhawanipatna , Odisha, India. E-
mail: dillip678in@yahoo.co.in
DOI: 10.9756/BIJAIP.10357
location decreases. Because of noise, low contrast and some
other factors, edge detection methods cannot give satisfactory
results, Marr and Hildreth [6]. There are some edge detection
algorithms in frequency domain, Musevi-Niya
andAghagolzadeh[7]where over or under edge based
segmentation occurs.
As the performance of classical edge detectors degrades
with noise, morphological edge detector has been studied, Lee
et al., [8].It was introduced by Matheron as a technique for
analyzing geometric structure of metallic solids. It was
extended to image analysis,Serra [9].Mathematical
morphology is a new mathematical theory which can be used
to process and analyze the images, Maragos [10],Richard [11]
and Jean [12]. It provides an alternative approach to image
processing based on shape,concept stemmed from set theory,
Serra [9]. In the mathematical morphology theory, images are
treated as sets, and morphological transformations which are
derived from Minkowski addition and subtraction are defined
to extract features in images. A fundamental advantage of
mathematical morphology applied to image is that it is
intuitive since it works directly on spatial domain.
The idea of fuzzy logic is to extend the binary (TRUE or
FALSE) computer model with some uncertainty or blur.Many
of our sensory impressions are qualitative and imprecise and,
therefore, unsuitable for accurate measurements. For example,
a pixel is perceived as “dark”, “bright” or even “very bright”,
but not as pixels with the gray scale value “231”. Fuzzy
quantities are based mathematically on the fuzzy set theory, in
which the belongingness of an element to a set of elements is
not restricted to the absolute states TRUE (1) or FALSE (0),
but continuously defined within the entire interval [0..1].
In general mathematical morphology, operation definitions
are similar to set theory and set operation definitions. For this
reason fuzzy set theory is easily applied to the mathematical
morphology, Nadarnejat [15]. Mathematical morphology is a
collection of operations which produces useful outcomes in
image processing area. It is completely based on set theory.
For this reason all of the operations in morphology are defined
on the simple set operation rules to apply them on image
pixels. The basic mathematical morphological operators are
dilation and erosion and the other morphological operations
are the synthesization of the two basic operations.The
objective of this paper is to present the hybrid approach for
edge detection. Under this technique, edge detection is
performed in two phases. In first phase, sobel algorithm is
applied for image smoothing and in second phase fuzzy
morphology is applied to detect actual edges. Fuzzy
morphology is a wonderful tool for edge detection. Fuzzy
mathematical morphology is beneficial for detection of the
Edge Detection Using Fuzzy Double Gradient
Morphology
Dillip Ranjan Nayak
T
ISSN: 2277-503X| © 2014 Bonfring