1
Canny Edge Detection {#tutorial_py_canny}
7
In this chapter, we will learn about
9
- Concept of Canny edge detection
10
- OpenCV functions for that : **cv2.Canny()**
15
Canny Edge Detection is a popular edge detection algorithm. It was developed by John F. Canny in
16
1986. It is a multi-stage algorithm and we will go through each stages.
18
-# **Noise Reduction**
20
Since edge detection is susceptible to noise in the image, first step is to remove the noise in the
21
image with a 5x5 Gaussian filter. We have already seen this in previous chapters.
23
-# **Finding Intensity Gradient of the Image**
25
Smoothened image is then filtered with a Sobel kernel in both horizontal and vertical direction to
26
get first derivative in horizontal direction (\f$G_x\f$) and vertical direction (\f$G_y\f$). From these two
27
images, we can find edge gradient and direction for each pixel as follows:
30
Edge\_Gradient \; (G) = \sqrt{G_x^2 + G_y^2} \\
31
Angle \; (\theta) = \tan^{-1} \bigg(\frac{G_y}{G_x}\bigg)
34
Gradient direction is always perpendicular to edges. It is rounded to one of four angles
35
representing vertical, horizontal and two diagonal directions.
37
-# **Non-maximum Suppression**
39
After getting gradient magnitude and direction, a full scan of image is done to remove any unwanted
40
pixels which may not constitute the edge. For this, at every pixel, pixel is checked if it is a
41
local maximum in its neighborhood in the direction of gradient. Check the image below:
43
![image](images/nms.jpg)
45
Point A is on the edge ( in vertical direction). Gradient direction is normal to the edge. Point B
46
and C are in gradient directions. So point A is checked with point B and C to see if it forms a
47
local maximum. If so, it is considered for next stage, otherwise, it is suppressed ( put to zero).
49
In short, the result you get is a binary image with "thin edges".
51
-# **Hysteresis Thresholding**
53
This stage decides which are all edges are really edges and which are not. For this, we need two
54
threshold values, minVal and maxVal. Any edges with intensity gradient more than maxVal are sure to
55
be edges and those below minVal are sure to be non-edges, so discarded. Those who lie between these
56
two thresholds are classified edges or non-edges based on their connectivity. If they are connected
57
to "sure-edge" pixels, they are considered to be part of edges. Otherwise, they are also discarded.
60
![image](images/hysteresis.jpg)
62
The edge A is above the maxVal, so considered as "sure-edge". Although edge C is below maxVal, it is
63
connected to edge A, so that also considered as valid edge and we get that full curve. But edge B,
64
although it is above minVal and is in same region as that of edge C, it is not connected to any
65
"sure-edge", so that is discarded. So it is very important that we have to select minVal and maxVal
66
accordingly to get the correct result.
68
This stage also removes small pixels noises on the assumption that edges are long lines.
70
So what we finally get is strong edges in the image.
72
Canny Edge Detection in OpenCV
73
------------------------------
75
OpenCV puts all the above in single function, **cv2.Canny()**. We will see how to use it. First
76
argument is our input image. Second and third arguments are our minVal and maxVal respectively.
77
Third argument is aperture_size. It is the size of Sobel kernel used for find image gradients. By
78
default it is 3. Last argument is L2gradient which specifies the equation for finding gradient
79
magnitude. If it is True, it uses the equation mentioned above which is more accurate, otherwise it
80
uses this function: \f$Edge\_Gradient \; (G) = |G_x| + |G_y|\f$. By default, it is False.
84
from matplotlib import pyplot as plt
86
img = cv2.imread('messi5.jpg',0)
87
edges = cv2.Canny(img,100,200)
89
plt.subplot(121),plt.imshow(img,cmap = 'gray')
90
plt.title('Original Image'), plt.xticks([]), plt.yticks([])
91
plt.subplot(122),plt.imshow(edges,cmap = 'gray')
92
plt.title('Edge Image'), plt.xticks([]), plt.yticks([])
98
![image](images/canny1.jpg)
103
-# Canny edge detector at [Wikipedia](http://en.wikipedia.org/wiki/Canny_edge_detector)
104
-# [Canny Edge Detection Tutorial](http://dasl.mem.drexel.edu/alumni/bGreen/www.pages.drexel.edu/_weg22/can_tut.html) by
110
-# Write a small application to find the Canny edge detection whose threshold values can be varied
111
using two trackbars. This way, you can understand the effect of threshold values.