av免费观看在线,免费岛国片,欧美高清一区二区

區域生長是一種串行區域分割的圖像分割方法。區域生長是指從某個像素出發，按照一定的準則，逐步加入鄰近像素，當滿足一定的條件時，區域生長終止。區域生長的好壞決定于1.初始點（種子點）的選取。2.生長準則。3.終止條件。區域生長是從某個或者某些像素點出發，最后得到整個區域，進而實現目標的提取。

區域生長的原理：

區域生長的基本思想是將具有相似性質的像素集合起來構成區域。具體先對每個需要分割的區域找一個種子像素作為生長起點，然后將種子像素和周圍鄰域中與種子像素有相同或相似性質的像素（根據某種事先確定的生長或相似準則來判定）合并到種子像素所在的區域中。將這些新像素當作新的種子繼續上面的過程，直到沒有滿足條件的像素可被包括進來。這樣一個區域就生長成了。

區域生長實現的步驟如下：

1. 對圖像順序掃描!找到第1個還沒有歸屬的像素, 設該像素為(x0, y0);

2. 以(x0, y0)為中心, 考慮(x0, y0)的4鄰域像素(x, y)如果(x0, y0)滿足生長準則, 將(x, y)與(x0, y0)合并(在同一區域內), 同時將(x, y)壓入堆棧;

3. 從堆棧中取出一個像素, 把它當作(x0, y0)返回到步驟2;

4. 當堆棧為空時!返回到步驟1;

5. 重復步驟1 - 4直到圖像中的每個點都有歸屬時。生長結束。

Python實現

				?

									import numpy as np

									import cv2

									class Point(object):

									 def __init__(self,x,y):

									  self.x = x

									  self.y = y

									 def getX(self):

									  return self.x

									 def getY(self):

									  return self.y

									def getGrayDiff(img,currentPoint,tmpPoint):

									 return abs(int(img[currentPoint.x,currentPoint.y]) - int(img[tmpPoint.x,tmpPoint.y]))

									def selectConnects(p):

									 if p != 0:

									  connects = [Point(-1, -1), Point(0, -1), Point(1, -1), Point(1, 0), Point(1, 1), \

									     Point(0, 1), Point(-1, 1), Point(-1, 0)]

									 else:

									  connects = [ Point(0, -1), Point(1, 0),Point(0, 1), Point(-1, 0)]

									 return connects

									def regionGrow(img,seeds,thresh,p = 1):

									 height, weight = img.shape

									 seedMark = np.zeros(img.shape)

									 seedList = []

									 for seed in seeds:

									  seedList.append(seed)

									 label = 1

									 connects = selectConnects(p)

									 while(len(seedList)>0):

									  currentPoint = seedList.pop(0)

									  seedMark[currentPoint.x,currentPoint.y] = label

									  for i in range(8):

									   tmpX = currentPoint.x + connects[i].x

									   tmpY = currentPoint.y + connects[i].y

									   if tmpX < 0 or tmpY < 0 or tmpX >= height or tmpY >= weight:

									    continue

									   grayDiff = getGrayDiff(img,currentPoint,Point(tmpX,tmpY))

									   if grayDiff < thresh and seedMark[tmpX,tmpY] == 0:

									    seedMark[tmpX,tmpY] = label

									    seedList.append(Point(tmpX,tmpY))

									 return seedMark

									img = cv2.imread('lean.png',0)

									seeds = [Point(10,10),Point(82,150),Point(20,300)]

									binaryImg = regionGrow(img,seeds,10)

									cv2.imshow(' ',binaryImg)

									cv2.waitKey(0)