Digital Signal Processing Reference
In-Depth Information
10 meters 20 meters 40 meters
Fig. 2.
The results of edge detection and center calculation of different position
In order to test the validity of algorithm, this means can be adopted that averages
results of repeated measuring same image replacing result of single measuring. Table 1
shows the center coordinates of ten consecutive measurements for above three images.
Table 1.
The center coordinates of different position
Measuring
times
Coordinates(X
i
, Y
j
)
10 meters
20 meters
40 meters
1
(161,114)
(144,137)
(159,130)
2
(161,116)
(145,139)
(157,134)
3
(162,115)
(144,139)
(159,133)
4
(161,116)
(144,138)
(158,132)
5
(162,115)
(146,140)
(160,133)
6
(161,116)
(145,137)
(158,129)
7
(161,114)
(145,140)
(160,135)
8
(162,114)
(144,139)
(157,130)
9
(160,115)
(145,139)
(159,113)
10
(162,115)
(144,138)
(157,133)
Average
(161.3,115.0)
(144.5,138.6)
(158.4,132.2)
Seen from that, the results prove that the algorithm is effective and accurate. The
average value of 10 times of measurements is also given in Table 1. The following is
standard deviation
of X and Y calculated by Eq.(15).
σ
∑
(15)
Where n is measuring times, V
i
is standardized residual, Table 2. shows the results of
standard deviation gained by calculating above 3 sets of data. It also illuminates
that the Gradient Hough Transform algorithm is effective to apply to collimating
measurement.
