像素之间的基本关系

🎯Some basic relationship between pixels——像素之间的基本关系

Neighbors and Connectivity——领域与联通

Neighbors of a pixel——像素的领域

4-neighbor——4领域,即\(N_{4}(p)\)

Diagonal neighbors——对角领域,即\(N_{D}(p)\)

8-neighbor——8领域,即\(N_{8}(p)\)

Connectivity——连通

相邻仅考虑像素间的空间关系

连通:空间上相邻且像素灰度值相似

两个像素是否连通:

  • 是否接触(相邻)
  • 灰度值是否满足某个特定的相似准则:灰度值相等或同在一个灰度值集合中
三种连通

假设V为灰度值集合,V ={1}:

  • 4-连通: 2个像素 p 和 r 在V 中取值且 r 在\(N_{4}(p)\)
  • 8-连通:2个像素 p 和 r 在V 中取值且 r 在\(N_{8}(p)\)
  • m-连通(混合连通):2个像素 p 和 r 在V 中取值,且满足下列条件之一
    ① r 在\(N_{4}(p)\)
    ② r 在\(N_{D}(p)\)中且集合\(N_{4}(p)\)\(N_{4}(r)\)是空集

Path——通路

 像素\(p(x, y)\)到像素\(q(s, t)\)的一条通路由一系列具有坐标\((x_0, y_0)\), \((x_1, y_1)\) ,…,\((x_i , y_i)\) ,…,\((x_n , y_n)\)的独立像素组成。这里 \((x, y)\)= \((x_0, y_0)\), \((x_n , y_n)\)= \((s, t)\),且 \((x_i , y_i)\)\((x_{i-1}, y_{i-1})\)连通。其中\(1≤i≤n\)\(n\)为通路长度。

通路种类:4-通路(4-path);8-通路(8-path);m-通路(m-path)

例题:

Distance Measures——距离度量

Definition:For pixels \(p, q, z\), with coordinates \((x, y), (s, t), (v, w)\) ,respectively, if

给出三个像素\(p, q, z\),坐标分别为\((x, y), (s, t), (v, w)\),则

  • \(D(p, q)≥0\) [\(D(p, q)= 0\), 当且仅当 p=q]
  • \(D(p, q)=D(q, p)\)
  • \(D(p, z)≤D(p, q) + D(q, z)\)

then D is a distance function or metric.

如过D满足以上三个条件,则称D为距离函数或距离度量

⚠️D距离与像素的点坐标相关
⚠️D距离与像素间的连通性无关

🎯Euclidean distance De——欧式距离(默认使用的距离函数)

Definition:

已知\(p(x,y)\),\(q(s,t)\)

\[D_e(p,q)=\left((x-s)^2+(y-t)^2\right)^{\frac{1}{2}} \]

For this distance measure, the pixels having a distance less than or equal to some value \(r\) from \((x,y)\) are the points contained in a disk of radius \(r\) centered at \((x,y)\).

距离点\((x,y)\)\(r\)距离的点组成一个圆

\(D_4\) distance (also called city-block distance)——\(D_4\)距离,即街区距离

\[D_4(p,q) = |x-s| + |y-t| \]

The pixels having a \(D_4\) distance from \((x, y)\) less than or equal to some value r form a diamond centered at \((x, y)\) .

The Pixels with \(D_4=1\) are the \(N_4\) of \((x, y)\).

\(D_8\) distance (chessboard distance)——\(D_8\)距离,即棋盘距离

\[D_8(p,q) = max(|x-s|,|y-t|) \]

The pixels with \(D_8\) distance from\((x, y)\) less than or equal to some value r form a square centered at \((x, y)\).

The pixels with \(D_8=1\) are the N8 of \((x, y)\).

下一篇:颜色基础与颜色模型

posted @ 2021-04-10 22:06  VanGy  阅读(828)  评论(0编辑  收藏  举报