Diyala Journal ISSN 1999-8716

of Engineering Printed in Iraq

Sciences
Second Engineering Scientific Conference
College of Engineering –University of Diyala
16-17 December. 2015, pp. 430-440

EDGE DETECTION-APPLICATION OF (FIRST AND

SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

Abdulbasit Alazzawi 1, Husam Alsaadi 2, Abidaoun Shallal 3, Saad Albwi 3

1
Science College, Diyala University, Iraq
2
Basic Education Colleges, Al-Mustansiriyah University, Iraq.
3
College of Engineering, Diyala University, Iraq.

ABSTRACT: - Edge detection is one of the most frequently used techniques in digital
image processing. Edges typically occur on the boundary between two different regions in an
image. In this paper the first method we will find the edge for image by using (1st Order
Derivative Filter ) method. In this method we take the 1st derivative of the intensity value
across the image and find points where the derivative is maximum then the edge could be
located. The gradient is a vector, whose components measure how rapid pixel value are
changing with distance in the x and y direction. In second method we use the (2nd Order
Derivative Operators) .
The 2nd derivative of an image where the image highlights regions of rapid intensity
change and is therefore often used for edge detection zero crossing edge detectors. The zero
crossing detector looks for places in the Laplacian of an image where the value of the
Laplacian passes through zero i.e. points where the Laplacian changes sign. The criteria that
used to comparison of results is (Root mean square error ) for different threshold values that
explain the result obtain by 2nd order are best of the result obtain by 1st order for all values of
threshold.
1. INTRODUCTION
The goals of edge detection are: Produce a line drawing of a scene from an image of
that scene, Important features can be extracted from the edges of an image (e.g. corners, lines
and curve) and These features are used by higher-level computer vision algorithms (e.g.,
recognition).Various physical events cause intensity changes:(a. Geometric events such as
Object boundary where discontinuity in depth and/or surface color and texture and Surface
boundary where discontinuity in surface orientation and/or surface color and texture).
(b. Non-geometric events such as Specularity where direct reflection of light, i.e. a mirror,
Shadows (from other objects or from the same object and Inter-reflections). The edge
detection can be found by four steps:-
1- Smoothing: suppress as much noise as possible, without destroying the true edges.
2- Enhancement: apply a filter to enhance the quality of the edges in the image (sharpening).
3- Detection: determine which edge pixels should be discarded as noise and which should be
retained (usually, threshold provides the criterion used for detection).
4- Localization: determine the exact location of an edge (sub-pixel resolution
might be required for some applications, that is, estimate the location of an edge to better
than the spacing between pixels). Edge thinning and linking are usually required here.
The Criteria for good edge filters depend on some factor as following:-
i-No response to flat regions ⇒ Sum of mask values is zero: ∑(r, c) = 0.
ii- Isotropy: Response must be independent of edge orientation.
iii- Good detection: Minimize the probabilities of (detecting spurious edges caused by noise
and missing real edges)
430
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

iv- Good localization: Detected edges must be as close as possible to true edges. v- Single
response: Minimize number of false local maxima around true edge [1].

2. METHODS OF EDGE DETECTION

The majority of different methods may be grouped into two categories
Gradient method: The gradient method detects the edges by looking for the maximum and
minimum in the first derivative of the image.
Laplacian method: It searches for zero crossings in the second derivative of the image to find
edges. An edge has the one-dimensional shape of a ramp and calculating the derivative of the
image can highlight its location. Suppose we have the following signal as shown in figure (1),
with an edge by the jump in intensity. If we take the gradient of this signal (which, in one
dimension, is just the first derivative with respect to t) we get the following signal as shown
in figure (2).
Clearly, the derivative shows a maximum located at the center of the edge in the
original signal. This method of locating an edge is characteristic of the “gradient filter”
family of edge detection filters and includes the Sobel method. A pixel location is declared an
edge location if the value of the gradient exceeds some threshold. As mentioned before, edges
will have higher pixel intensity values than those surrounding it. So once a threshold is set,
you can compare the gradient value to the threshold value and detect an edge whenever the
threshold is exceeded. Furthermore, when the first derivative is at a maximum, the second
derivative is zero. As a result, another alternative to finding the location of an edge is to
locate the zeros in the second derivative. This method is known as the Laplacian method and
the second derivative of the signal is shown in figure (3) [2].
2-1. 1st Order Derivative Filter (Sobel Filter)
Most edge detection methods work on the assumption that the edge occurs where
there is a discontinuity in the intensity function or a very steep intensity gradient in the
image. Using this assumption, if one take the derivative of the intensity value across the
image and find points where the derivative is maximum then the edge could be located. The
gradient is a vector, whose components measure how rapid pixel value are changing with
distance in the x and y direction. Thus, the components of the gradient found by using the
following equations (1) & (2).

𝜕𝑓(𝑥, 𝑦) 𝑓 (𝑥 + 𝑑𝑥, 𝑦) − 𝑓(𝑥, 𝑦)

= ∆𝑥 = … … . (1)
𝜕𝑥 𝑑𝑥

𝜕𝑓(𝑥, 𝑦) 𝑓 (𝑥, 𝑦 + 𝑑𝑦) − 𝑓(𝑥, 𝑦)

= ∆𝑦 = … … . (2)
𝜕𝑦 𝑑𝑦

Where dx & dy measure distance along the x and y directions respectively. In discrete
images, one can consider dx & dy in terms of numbers of pixel between two points. dx = dy =
1 (pixel spacing) is the point at which pixel coordinates are(i, j) thus, the value of
(∆𝑥 𝑎𝑛𝑑 ∆𝑦) can calculated by equations (3) & (4).

∆𝑥 = 𝑓 (𝑖 + 1, 𝑗) − 𝑓(𝑖, 𝑗) … … … (3)

∆𝑦 = 𝑓 (𝑖, 𝑗 + 1) − 𝑓 (𝑖, 𝑗) … … … (4)

In order to detect the presence of a gradient discontinuity, one could calculate the
change in the gradient at (i, j) .This can be done by finding the following magnitude measure
and the gradient direction θ is given by the equation (5).

Diyala Journal of Engineering Sciences, Vol. 08, No. 04, Special Issue
431
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING
∆𝑦
∅ = tan−1 [ ] … … . . (5)
∆

The Sobel operator is an example of the gradient method. It is a discrete

differentiation operator, computing an approximation of the gradient of the image intensity
function [3].

−1 1 −1 0
∆𝑥 = [ ], ∆𝑦 = [ ]
0 0 1 0

An advantage of using a larger mask size is that the errors due to the effects of noise
are reduced by local averaging within the neighborhood of the mask. An advantage of using a
mask of odd size is that the operators are centered and can therefore provide an estimate that
is based on a center pixel (i,j). One important edge operator of this type is the Sobel edge
operator. The Sobel edge operator masks are given as:

−1 0 1 1 2 1
[
∆𝑥 = −2 0 2], [
∆𝑦 = 0 0 0]
−1 0 1 −1 −1 −1

The operator calculates the gradient of the image intensity at each point, giving the
direction of the largest possible increase from light to dark and the rate of change in that
direction. The result therefore shows how "abruptly" or "smoothly" the image changes at that
point and therefore how likely it is that part of the image represents an edge, as well as how
that the edge is likely to be oriented. In practice, the magnitude (likelihood of an edge)
calculation is more reliable and easier to interpret than the direction calculation.
Mathematically, the gradient of a two-variable function (the image intensity function) at each
image point is a 2D vector with the components given by the derivatives in the horizontal and
vertical directions. At each image point, the gradient vector points to the direction of largest
possible intensity increase, and the length of the gradient vector corresponds to the rate of
change in that direction. This implies that the result of the Sobel operator at any image point
which is in a region of constant image intensity is a zero vector and at a point on an edge is a
vector which points across the edge, from darker to brighter values [3].

2-2. 2nd Order Derivative Operators (Laplacian Filter)

The Laplacian is a 2-D measure of the 2nd derivative of an image. The Laplacian of
an image highlights regions of rapid intensity change and is therefore often used for edge
detection zero crossing edge detectors). The Laplacian is often applied to an image that has
first been smoothed with something approximating a Gaussian smoothing filter in order to
reduce its sensitivity to noise. The operator normally takes a single gray level image as input
and produces another binary image as output. The zero crossing detector looks for places in
the Laplacian of an image where the value of the Laplacian passes through zero i.e. points
where the Laplacian changes sign. Such points often occur at edges in images i.e. points
where the intensity of the image changes rapidly, but they also occur at places that are not as
easy to associate with edges. It is best to think of the zero crossing detector as some sort of
feature detector rather than as a specific edge detector. Zero crossings always lie on closed
contours, and so the output from the zero crossing detectors is usually a binary image with
single pixel thickness lines showing the positions of the zero crossing points [4,5].The
derivative operator Laplacian for an Image is defined as shown in equations (6), (7) and (8):

𝜕2𝑓 𝜕2𝑓
∆2 𝑓 = + … . … . . (6)
𝜕𝑥 2 𝜕𝑦 2
Diyala Journal of Engineering Sciences, Special Issue
432
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING
𝜕2𝑓
For X-direction, = 𝑓 (𝑥 + 1, 𝑦) + 𝑓(𝑥 − 1, 𝑦) − 2𝑓(𝑥, 𝑦) … … . . (7)
𝜕𝑥 2

𝜕2𝑓
For Y-direction, = 𝑓 (𝑥, 𝑦 + 1) + 𝑓(𝑥, 𝑦 − 1) − 2𝑓(𝑥, 𝑦) … … . . (8)
𝜕𝑦 2

By substituting, Equations (7) and (8) in (6), we obtain the equation (9)

∆2 𝑓 (𝑥, 𝑦) = 𝑓 (𝑥 + 1, 𝑦) + 𝑓 (𝑥 − 1, 𝑦) + 𝑓 (𝑥, 𝑦 + 1) + 𝑓(𝑥, 𝑦 − 1) − 4𝑓(𝑥, 𝑦) … (9)

If we use the value of Mask as explain blow we obtain on equation (9):

0 1 0
1 -4 1
0 1 0

3. THRESHOLD AND THINNING

i. Threshold
Once we have computed a measure of edge strength (typically the gradient
magnitude), the next stage is to apply a threshold, to decide whether edges are present or not
at an image point. The lower the threshold, the more edges will be detected, and the result
will be increasingly susceptible to noise and detecting edges of irrelevant features in the
image. Conversely a high threshold may miss subtle edges, or result in fragmented edges. If
the edge thresholding is applied to just the gradient magnitude image, the resulting edges will
in general be thick and some type of edge thinning post-processing is necessary. For edges
detected with non-maximum suppression however, the edge curves are thin by definition and
the edge pixels can be linked into edge polygon by an edge linking (edge tracking) procedure.
On a discrete grid, the non-maximum suppression stage can be implemented by estimating
the gradient direction using first-order derivatives, then rounding off the gradient direction to
multiples of 45 degrees, and finally comparing the values of the gradient magnitude in the
estimated gradient direction. We begin by using the upper threshold to find the start of an
edge. Once we have a start point, we then trace the path of the edge through the image pixel
by pixel, marking an edge whenever we are above the lower threshold. We stop marking our
edge only when the value falls below our lower threshold. This approach makes the
assumption that edges are likely to be in continuous curves, and allows us to follow a faint
section of an edge we have previously seen, without meaning that every noisy pixel in the
image is marked down as an edge. Still, however, we have the problem of choosing
appropriate thresholding parameters, and suitable thresholding values may vary over the
image. [8, 9]

ii. Thinning
Edge thinning is a technique used to remove the unwanted spurious points on the
edges in an image. This technique is employed after the image has been filtered for noise
(using median, Gaussian filter etc.), the edge operator has been applied to detect the edges
and after the edges have been smoothed using an appropriate threshold value. This removes
all the unwanted points and if applied carefully, results in one pixel thick edge elements. It
can be used for several applications, but is particularly useful for skeletonization. In this
mode it is commonly used to tidy up the output of edge detectors by reducing all lines to
single pixel thickness. Thinning is normally only applied to binary images, and produces
another binary image as output. The thinning operation is related to the hit-and-miss
transform, and so it is helpful to have an understanding of that operator before reading on.

Diyala Journal of Engineering Sciences, Vol. 08, No. 04, Special Issue
433
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

4. Result and Conclusion

The results that have been obtained by using )256 * 256( gray scale image as shown
in Figure (4), after applied the first order derivative (Sobel filter) and second order derivative
(laplacian filter) we gate on (gradient magnitude, gradient direction, horizontal gradient X-
axis and vertical gradient Y-axis) as shown in figures (5) to (8) .But only the gradient
magnitude use for determine edge, and compare with edge obtained by MATLAB function
(canny filter) as shown in figure (9) to find root mean square error. The figures from (10) to
(19) shows the result for different threshold (from 50 to 90), where the threshold here
represent the value of gray level. For each threshold we find the root mean square error that
represents the truncation error. The root mean square is calculate between the result that
obtain by applied the canny filter as true result and the results that obtain from (1 st & 2nd )
order as numerical result .The value of root mean square errors used as Objective fidelity
criteria as shown in equation (10). Another fidelity criteria is subjective that performed by
gathering a group of people that are representative the desired population, and then having
all the test subjects evaluate the result according to a predefine scoring criterion. From result
we note the quality of edges are increase with increase the value of threshold and the values
of root mean square error are decrease as shown in table(1). This mean the information loses
in first order more than the information loses in second order.
𝑁−1
𝑁−1
rmse=sqrut (1/N²∑ ∑𝑐=0 [𝑓(𝑟, 𝑐 ) − ℎ(𝑟, 𝑐)]² …….. (10)
𝑟=0
Where the 𝑓 (r,c) is edge obtain by applied canny filter, and ℎ (r,c) is the edge obtain by applied
1st and 2nd order derivative.

REFERENCE
1) Edge-Detection-Trucco, Chapt-4-AND-Jain-et-al-Chapt-5-Definition of Edges, Edges are
significant local changes of intensity.
2) Edge Detection Tutorial. Author: Bill Green, (2002).
3) A Descriptive Algorithm for Sobel Image Edge Detection, O. R. Vincent and O.
Folorunso, Proceedings of Informing Science & IT Education Conference (InSITE),
(2009).
4) Filtering Corrupted Image and Edge Detection in Restored Gray scale Image Using
Derivative Filters, Chandra Sekhar Panda,Prof. (Dr.) Srikanta Patnaik, International
Journal of Image Processing, (2009).
5) Digital Image Processing (3rd Edition) Hardcover, by Rafael C. Gonzalez & Richard,
(2007).
6) Numerical Methods for Engineers and Scientists, An Introduction with Applications using
MATLAB, Third Edition, Amos Gilat & Vish Subramaniam, (2014).
7) Digital Image Processing Algorithms and Application, by Ioannis Pitas, (2001).
8) Shapiro Computer and Robot Vision, Addison-Wesley Publishing Company, R. Haralick
and L., 1992.
9) E. Davies Machine Vision: Theory, Algorithms and Practicalities, Academic Press, 1990.
10) E. Gose, R. Johnsonbaug, S. Jost, Pattern Recognition and Image Analysis,
a. Prentice-Hall, Englewood Cliffs, NJ, 1996
11) A. Cumani, “Edge detection in multispectral images,” CVGIP: Graphical
a. Models and Image Proc. 1991.
b. D. Marr and E. Hildreth, "Theory of Edge Detection," Proceedings of the Royal Society
of London, B207, 1980.

Diyala Journal of Engineering Sciences, Special Issue

434
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

Table (1): Show how the Value of Root Mean Square Error for (1st and 2nd) Derivative
Change with Different Value of Threshold.
No Threshold Value Rmse for 1st derivtive Rmse for 2nd derivtive
1 50 0.59743 0.51934
2 60 0.57274 0.49411
3 70 0.54882 0.54882
4 80 0.53078 0.4573
5 90 0.51542 0.44253

Figure (1): Signal Curve. Figure (2): Max Value Derivative

Figure (3): Zero Crossing Derivative

original image size (256*256)

Figure (4): The Original Image with Size (256 * 256)

Diyala Journal of Engineering Sciences, Vol. 08, No. 04, Special Issue
435
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

Gradient magnitude 1st derivative Gradient magnitude 2nd derivative

Figure (5): Gradient Magnitude for (1st and 2nd) Derivative

Gradient direction 1st derivative Gradient direction 2nd derivative

Figure (6): Gradient Direction for (1st and 2nd) Derivative

Horizintal gradient 1st derivative : X axis Horizintal gradient 2nd derivative: X axis

Figure (7): Horizontal Gradient for (1st and 2nd Derivative

Diyala Journal of Engineering Sciences, Special Issue

436
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

Vertival gradient 1st derivative: Y axis Vertival gradient 2nd derivative: Y axis

Figure (8): Vertical Gradient for (1st and 2nd) Derivative

Egde obtain by MatLab Function

Figure (9): Edge obtain by matlab function

1st order After Threshold 2nd order After Threshold

Figure (10): Edge of Image for (1st and 2nd) Derivative with Threshold=50
Diyala Journal of Engineering Sciences, Vol. 08, No. 04, Special Issue
437
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

1st order After Thinning 2nd order After Thinning

Figure (11): Edge after Thinning for (1st and 2nd) Derivative with Threshold=50

1st order After Threshold 2nd order After Threshold

Figure (12): Edge after Threshold for (1st and 2nd) Derivative with Threshold=60

1st order After Thinning 2nd order After Thinning

Figure (13): Edge after Thinning for (1st and 2nd) Derivative with Threshold=60

Diyala Journal of Engineering Sciences, Special Issue

438
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

1st order After Threshold 2nd order After Threshold

Figure (14): Edge after Threshold for (1st and 2nd) Derivative with Threshold=70

1st order After Thinning 2nd order After Thinning

Figure (15): Edge after Thinning for (1st and 2nd) Derivative with Threshold=70

1st order After Threshold 2nd order After Threshold

Figure (16): Edge after Threshold for (1st and 2nd) Derivative with Threshold=80

Diyala Journal of Engineering Sciences, Vol. 08, No. 04, Special Issue
439
Second Engineering Scientific Conference-College of Engineering –University of Diyala 16-17 December. 2015
EDGE DETECTION-APPLICATION OF (FIRST AND SECOND) ORDER DERIVATIVE IN IMAGE PROCESSING

1st order After Thinning 2nd order After Thinning

Figure (17): Edge after Thinning for (1st and 2nd) Derivative with Threshold=80

1st order After Threshold 2nd order After Threshold

Figure (18): Edge after Threshold for (1st and 2nd) Derivative with Threshold=90

1st order After Thinning 2nd order After Thinning

Figure (19): Edge after Thinning for (1st and 2nd) Derivative with Threshold=90

Diyala Journal of Engineering Sciences, Special Issue

440