4.3 RFID Tag Positioning Method
131
of the algorithm is to determine the features that are easy to distinguish and identify
and determine the matching position through the similarity of each feature set.
4.3.2 Image Threshold Segmentation Technology
Determining the characteristics of the tag and separating the tag from the background
are important contents of tag positioning. If there is a significant difference between
the tag background and the grayscale set of the tag, and the two grayscale sets can
be segmented by the threshold, then the image can be segmented by the threshold to
find the tag in the image.
For example, if a method of threshold image segmentation is selected, a corresponding threshold value needs to be preset in conjunction with the gray value of
the tag, and the number of threshold values may be one or more. Usually combined
with the display requirements, the target and background are represented by 0 and
255, respectively. Because the background and the target may not only exist in the
two gray sets, it is necessary to set several thresholds during the process of extracting
the target. If the value of the pixel is within the threshold, then the system will be
recognized as the target, and on the contrary, the system recognized as a background
can be described by the following formula:
g(x, y) =
Z E , T 1 ≤ f (x, y) ≤ T 2
Z B , Othercases
(4.24)
Among them, Z E and Z B represent the grayscale of the target and background in
the image, respectively, and [T 1 , T 2 ] is the selected threshold range.
For threshold segmentation, if the extraction target or background has only
small grayscale fluctuations, then better extraction effects can be exerted. Only the
threshold value needs to be understood as a fixed value. In such a case, only setting
a reasonable threshold can achieve the purpose of image segmentation. However,
the target and background usually have large grayscale fluctuations, especially if
the environment is more complex, then the grayscale of the target and part of the
background may be very close, so part of the background content may be judged as
the target, and the problem of misidentification may occur, which affects subsequent
processing. In such cases, it is necessary to set the corresponding threshold value
in accordance with the difference in the image position. There are several common
methods to choose the best threshold:
(1) Maximum variance threshold
The basic idea of the maximum variance threshold is to set a certain threshold and
divide the histogram to obtain two groups. The method for determining the threshold
is to make the two groups have the largest variance. A histogram is a method of
expressing an image in a statistical manner. A series of vertical stripes with different
131
of the algorithm is to determine the features that are easy to distinguish and identify
and determine the matching position through the similarity of each feature set.
4.3.2 Image Threshold Segmentation Technology
Determining the characteristics of the tag and separating the tag from the background
are important contents of tag positioning. If there is a significant difference between
the tag background and the grayscale set of the tag, and the two grayscale sets can
be segmented by the threshold, then the image can be segmented by the threshold to
find the tag in the image.
For example, if a method of threshold image segmentation is selected, a corresponding threshold value needs to be preset in conjunction with the gray value of
the tag, and the number of threshold values may be one or more. Usually combined
with the display requirements, the target and background are represented by 0 and
255, respectively. Because the background and the target may not only exist in the
two gray sets, it is necessary to set several thresholds during the process of extracting
the target. If the value of the pixel is within the threshold, then the system will be
recognized as the target, and on the contrary, the system recognized as a background
can be described by the following formula:
g(x, y) =
Z E , T 1 ≤ f (x, y) ≤ T 2
Z B , Othercases
(4.24)
Among them, Z E and Z B represent the grayscale of the target and background in
the image, respectively, and [T 1 , T 2 ] is the selected threshold range.
For threshold segmentation, if the extraction target or background has only
small grayscale fluctuations, then better extraction effects can be exerted. Only the
threshold value needs to be understood as a fixed value. In such a case, only setting
a reasonable threshold can achieve the purpose of image segmentation. However,
the target and background usually have large grayscale fluctuations, especially if
the environment is more complex, then the grayscale of the target and part of the
background may be very close, so part of the background content may be judged as
the target, and the problem of misidentification may occur, which affects subsequent
processing. In such cases, it is necessary to set the corresponding threshold value
in accordance with the difference in the image position. There are several common
methods to choose the best threshold:
(1) Maximum variance threshold
The basic idea of the maximum variance threshold is to set a certain threshold and
divide the histogram to obtain two groups. The method for determining the threshold
is to make the two groups have the largest variance. A histogram is a method of
expressing an image in a statistical manner. A series of vertical stripes with different