PasteAbstract:- Interpretation of image contents is one of the main objectives in computer vision. Edge is a basic and important feature of an image. Image is a combination of edges. Edges of an image are considered a type of crucial information that can be extracted by applying detectors with different methodology. Edge detection is a vital step as it is a process of identifying and locating sharp discontinuities in an image. Gravitational Search Algorithm (GSA) is a new population based search algorithm inspired by Newtonian gravity. In this paper, comprehensive analysis is been done on Sobel edge detection method and Gravitational search algorithm. Experimental analysis are done on several test images by using these two methods which shows that GSA method has superior image quality and better Structural Similarity Index Metric (SSIM) value.
1. INTRODUCTION:-
Today, with the development of image processing technology it is used quite a lot. Most of our lives intertwined with a technological structure is wrapped in an image processing device is configured on. The algorithms used in this field of technology with the speed proportional manner are developing very rapidly. One of the most important areas in the image processing is an edge detection process. Edge detection is the important basis of image analysis, including image segmentation, texture feature extraction and shape feature extraction [1]. Edge detection is the process of identifying and locating sharp discontinuities in an image. It is the process of locating an edge of an image. The main purpose of edge detection is to simplify the image data in order to minimize the amount of data to be processed. Edge detection can be done in four steps smoothing, enhancement, detection and localization. The objective of edge detection is to discover the information about the shapes and the reflectance or transmittance in an image and every detector is used to avoid false edges and detected edges should closest to reference edges [2].
Gravitational search algorithm is an optimization algorithm inspired by Newtonian gravity. Algorithm uses the theory of Newtonian gravity and its searcher agents are the collection of masses. Masses cooperate using a direct form of communication, through gravitational force of attraction. Each mass presents a solution, and the algorithm is navigated by properly adjusting the gravitational and inertia masses. The lighter masses tend to get attracted towards heaviest mass. The heavier mass presents an optimum solution in the search space [3].
2. LITERATURE SURVEY:-
1. Om Prakash Verma and Rishabh Sharma proposed a approach for edge detection using gravitational search algorithm and universal law of gravity. In this approach the edges are are detected using the law of gravity and movement of agents are computed using gravitational search algorithm.It leads to a minimal set of input data to be processed and making the process much faster and memory efficient [4].
2. Ghassan Mahmood husein Amer and Dr..Ahmed Mohamed Abushaala presented methods for edge segmentation of images.The five techniques was used and these techniques were compared to each other to choose the best technique for edge detection segment image [5].
3. Kokila Jandial, Nitika Kapoor, Harish Kundra proposed approach which was able to detect the edge pixel in an image. It also assumes that each image pixel is a celestial body with a mass represented by its gray scale intensity. This approach was focused on optimization of edge detection problem [6].
4. Yusra A.Y.Al-Najjar and Dr. Den Chen Soong compared various techniques for measuring the quality of the image such as PSNR, HVS, SSIM, UIQI. This presents that the good IQM must be accurate and consistent in predicting the quality [7].
3. EDGE DETECTION TECHNIQUES:-
Edge Detection is the name for a set of mathematical methods which aim at identifying points in a digital image at which the image brightness changes sharply or, more formally has discontinuities. The points at which image brightness changes sharply are typically organized into a set of curved line segments termed edges. Edge detection can be defined as a process to locate the edges that has good orientation and it is an essential tool of image segmentation. Edge detection method transforms the original image into edge image with the help of operators. In the process of edge detection, the image is given as input image and then that image converts into gray scale image. And then apply the edge detector to detect and extract the edges present within an image as output [8].
The edge detection algorithms can be generally classified based on the behavioral study of edges with respect to the operators. Different edge-detection approaches can be broadly classified under Classical or Gradient based edge detectors (first derivative), Zero crossing (second derivative) and Optimal edge-detector [9].
Methods Approaches
First order derivative/ Gradient methods Sobel Operator
Prewitt Operator
Roberts Operator
Second order derivative/ Zero crossing Laplacian of Gaussian
Difference of Gaussian
Optimal Edge Detection Canny Edge Detector
Table:-1 Edge Detection Approaches
3.1 Sobel Edge Detection:- Irwin Sobel has proposed the Sobel edge detection technique in 1970. The Sobel operator performs a 2-D spatial gradient measurement on an image and so emphasizes regions of high spatial frequency that correspond to edges. The Sobel edge detector uses two masks, one vertical and one horizontal.
These masks are generally used 3×3 matrices. Especially, the matrices which have 3×3 dimensions are used in matlab. Sobel has two main advantages:-
(i) Since the introduction of the average factor, it has smoothing effect to the random noise of the image.
(ii) Because it is the differential of two rows or two columns, so the elements of the edge on both sides has been enhanced, so that the edges seems thick and bright [10].
One kernel is simply the other rotated by 90o. These kernels are designed to respond to edges running vertically and horizontally relative to the pixel grid, one kernel for each of the two perpendicular orientations. The kernels can be applied separately to input image to produce separate measurement of gradient component in each orientation which can be combined to find the absolute magnitude of gradient at each point [11]. The partial derivatives in x and y direction is given as follows:-
Sx={f(x+1,y-1)+2f(x+1,y)+f(x+1,y+1)}-{f(x-1,y-1)+2f(x-1,y)+f(x-1,y+1)} Sy={f(x-1,y+1)+2f(x, y)+f(x+1,y-1)}-{f(x-1,y-1)+2f(x, y)+f(x+1,y-1)}
The gradient of each pixel is calculated using:-
g(x, y) = √( 2 + 2)
-1 0 +1
-2 0 +2
-1 0 +1
Sx
+1 +2 +1
0 0 0
-1 -2 -1
Fig:-1 Masks used by the Sobel operator
Algorithm:-
Step 1: Read the image.
Step 2: Preprocessing the image.
Step 3: Perform masking , .
Step 4: Set the Threshold value.
Step 5: Evaluate the SSIM parameter with the edge map image.
4. GRAVITATIONAL SEARCH ALGORITHM:-
GSA was introduced by Rashedi et al. in 2009 and is intended to solve optimization problems. The population-based heuristic algorithm is based on the law of gravity and mass interactions. The algorithm is comprised of collection of searcher agents that interact with each other through the gravity force. The agents are considered as objects and their performance is measured by their masses. The gravity force causes a global movement where all objects move towards other objects with heavier masses. The masses are actually obeying the law of gravity as shown in Equation (1) and the law of motion in
Equation (2).
F = G (M1M2 / R2) (1)
a = F/M (2)
Based on Equation (1), F represents the magnitude of the gravitational force, G is gravitational constant, M1 and M2 are the mass of the first and second objects and R is distance between the two objects. Equation (1) shows that in the Newton law of gravity, the gravitational force between two objects is directly proportional to the product of their masses and inversely proportional to the square of the distance between the objects. While for Equation (2), Newton’s second law shows that when a force, F is applied to an object, its acceleration, a, depends on the force and its mass, M.
In GSA, the agent has four parameters which are position, inertial mass, active gravitational mass, and passive gravitational mass. The position of the mass represents the solution of the problem, where the gravitational and inertial masses are determined using a fitness function. The algorithm is navigated by adjusting the gravitational and inertia masses, whereas each mass presents a solution. Masses are attracted by the heaviest mass. Hence, the heaviest mass presents an optimum solution in the search space [12].
The GSA algorithm is mainly comprises of the following of the following steps:-
(i). Identification of search space.
(ii). Initialization.
(iii). Agent evaluation using fitness function.
(iv). Update G(t), best(t), worst(t) and ( )for i=1,2,….N.
(v). Calculation of the total force in all possible directions.
(vi). Acceleration and velocity calculations.
(vii). Updating agent’s position.
(viii). Repeat steps 3 to 7 until the stop criterion is reached.
(ix). End.
Generate Initial Population
Evaluate the fitness for each agent
Update the G, best and worst of the population
Calculate M and a for each agent
Update Velocity and Position
No
Meeting end of
criterion?
Yes
Return best solution
Fig:-2 Flowchart of GSA
Algorithm:-
Step 1: Read the image.
Step 2: Preprocessing the image.
Step 3: Evaluation using fitness function.
Step 4: Update G(t), best(t), worst(t) and M(t).
Step 5: Calculation of total force in all possible directions.
Step 6: Now, using adaptive Thresholding.
Step 7: Set the Threshold value.
Step 8: Evaluate the SSIM parameter with the edge map image.
5. Parameter Used for Comparison:-
In this paper, SSIM is used to measure the image quality of each edge of both methods. Structural Similarity Index Metric (SSIM):- The structural similarity index is a method used for measuring image quality. It is measuring of image quality based on an initial compressed or distortion-free image as reference. It computes the quality of a distorted image by computing the correlations in luminance, contrast and structure locally between the reference and distorted images and averaging these quantities over the entire image. It is an improved version of the universal image quality index. The SSIM index values vary between 0 and 1. The values close to 1 shows the highest correspondence with the original images [13].
The Structural Similarity (SSIM) Index quality assessment is based on the computation of three terms, namely the luminance term, the contrast term and the structural term. The overall index is a multiplicative combination of the three terms.
( , ) = [ ( , )] .[ ( , )] .[ ( , )]
6. EXPERIMENTAL ANALYSIS:-
Original Image Sobel Edge Map Image GSA Edge Map Image
Image1.jpg
Image2.jpg
Image3.jpg
Image4.jpg
Image5.png
Fig:-3 Result Images of Sobel & GSA
7. RESULT ANALYSIS:-
Image Sobel SSIM Value GSA SSIM Value
Image1.jpg 0.9839 0.9983
Image2.jpg 0.9775 0.9973
Image3.jpg 0.9709 0.9966
Image4.jpg 0.9672 0.9965
Image5.png 0.9639 0.9953
Table:-2 Comparison of Sobel & GSA based on SSIM parameter
8. CONCLUSION:-
Edge is a basic feature of an image. In this paper, the investigation has been done on image quality on a selected test images using Sobel operator and Gravitational Search Algorithm based on Structural Similarity Index (SSIM) parameter, which demonstrates that Gravitational Search Algorithm has a superior image quality and better Structural Similarity Index (SSIM) value.
your essay in here…