File tree Expand file tree Collapse file tree 5 files changed +5
-16
lines changed Expand file tree Collapse file tree 5 files changed +5
-16
lines changed Original file line number Diff line number Diff line change 32
32
33
33
** 2 高斯函数**
34
34
35
-
36
-
37
- ![ img] ( https:////upload-images.jianshu.io/upload_images/14512145-cb79bc3d41cc37fd.png?imageMogr2/auto-orient/strip%7CimageView2/2/w/708/format/webp )
38
-
39
-
35
+ ![ ] ( imgs/DLIB-0024.png )
40
36
41
37
注:σ的大小决定了高斯函数的宽度。
42
38
47
43
举个栗子:
48
44
假定中心点的坐标是(0,0),那么取距离它最近的8个点坐标,为了计算,需要设定σ的值。假定σ=1.5,则模糊半径为1的高斯模板就算如下
49
45
50
-
51
-
52
- ![ img] ( https:////upload-images.jianshu.io/upload_images/14512145-04d41990169b094a.png?imageMogr2/auto-orient/strip%7CimageView2/2/w/800/format/webp )
53
-
54
-
46
+ ![ ] ( imgs/DLIB-0025.png )
55
47
56
48
这个时候我们我们还要确保这九个点加起来为1(这个是高斯模板的特性),这9个点的权重总和等于0.4787147,因此上面9个值还要分别除以0.4787147,得到最终的高斯模板。
57
49
58
-
59
-
60
- ![ img] ( https:////upload-images.jianshu.io/upload_images/14512145-049ada57d888bf79.png?imageMogr2/auto-orient/strip%7CimageView2/2/w/409/format/webp )
50
+ ![ ] ( imgs/DLIB-0026.png )
61
51
62
52
** 4 高斯滤波计算**
63
53
64
54
有了高斯模板,那么高斯滤波的计算便顺风顺水了。
65
- 举个栗子:
66
- 假设现有9个像素点,灰度值(0-255)的高斯滤波计算如下:
55
+ 举个栗子:假设现有9个像素点,灰度值(0-255)的高斯滤波计算如下:
67
56
68
- ![ img ] ( https:////upload-images.jianshu.io/upload_images/14512145-530497b10b412a95. png?imageMogr2/auto-orient/strip%7CimageView2/2/w/1000/format/webp )
57
+ ![ ] ( imgs/DLIB-0027. png)
69
58
70
59
参考来源:(https://blog.csdn.net/nima1994/article/details/79776802)
71
60
You can’t perform that action at this time.
0 commit comments