Edge filters can used emphasize these rapid changes gray value while
also smoothing areas with more subtle changes gray value. does not smooth equally well all directions (i.
Average value filter: With the simple average filter value (also called box filter), all pixels are weighted equally,
which also means that the filter anisotropic, i.e.
Gain 1/16
1 1
2 2
1 1
Offset 0
Edge filters
Edges represent large jumps the gray value range image and provide basic information about the
shape and contour objects.
>>>
34 Smart Camera Smart Sensor User's manual 1.4.23
.e.
Gain 1/9
1 1
1 1
1 1
Offset 0
Binomial filter: contrast the simple average filter, the binomial filter works equally well all directions
with regard smoothing and noise reduction. following the coefficients the binomial filter, noise minimized and edges are emphasized.
Laplace: common filter for edge detection the Laplace filter, which subtracts smoothed image from
the original image.
Gain 1/8
-1 -1
-1 -1
-1 -1
Offset 127
Sobel filter: The Sobel filter, however, needs two internal filter masks work the horizontal and vertical
direction. the
corner pixels furthest away from the center pixel). The resulting image the same the original image.
Gain 1
0 0
0 0
0 0
Offset 0
Smoothing filters
Smoothing filters reduce image noise but also all soften other structures, such edges.
Gain 1/4
SobelX
1 -1
2 -2
1 -1
Offset 0
Gain 1/4
SobelY
1 1
0 0
-1 -1
Offset 0
Sobel abs(SobelX) abs(SobelY)
Example: Sobel filtering shot emphasize the edges.1 Common filter types
Keep
Default setting for user-defined filters.Function description
7
