美文网首页
图像轮廓(2)

图像轮廓(2)

作者: Tsukinousag | 来源:发表于2021-11-03 22:23 被阅读0次

    1. 凸包

    凸缺陷可以用来处理手势识别问题

    Hull = cv2.convexHull(points,clockwise,returnPoints)

    points：轮廓
    clockwise：True:凸包按顺时针方向排列；False：凸包按逆时针排列
    returnPoints：True：返回凸包角点（x,y）坐标 ；False：返回轮廓中凸包角点的索引

    import cv2

img=cv2.imread("D:/part12/part12_8.bmp")
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

hull1=cv2.convexHull(contours[0])
print("默认是False,hull的值",hull1)

hull2=cv2.convexHull(contours[0],returnPoints=False)
print("是True,hull的值",hull2)

    默认是False,hull的值 [[[256  28]]

 [[260  29]]

 [[319  45]]

 [[392  91]]

...
是True,hull的值 [[526]
 [524]
 [428]
 [347]
...

    绘制凸包

    import cv2

img=cv2.imread("D:/part12/part12_8.bmp")
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

hull=cv2.convexHull(contours[0])

cv2.polylines(img,[hull],True,(0,255,0),2)

cv2.imshow("img",img)

cv2.waitKey(0)
cv2.destroyWindow()

    2. 凸缺陷

    凸包与轮廓间的部分称为凸缺陷

    convexityDefects=cv2.convexityDefects(contour,convexhull)

    返回值：[起点，终点，轮廓上距离凸包最远的点，最远点的近似距离]

    前三个值是轮廓点的索引

    参数：contour是轮廓；convexhull是凸包

    因为用到的是轮廓点的索引，所以cv2.convexHull的参数returnPoints的值必须是False

    import cv2

img=cv2.imread("D:/part12/part12_8.bmp")
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

cnt=contours[0]

hull=cv2.convexHull(cnt,returnPoints=False)

defects=cv2.convexityDefects(cnt,hull)

print("defects:\n",defects)

##构造凸缺陷
for i in range(defects.shape[0]):
    s,e,f,d=defects[i,0]
    start=tuple(cnt[s][0])
    end=tuple(cnt[e][0])
    far=tuple(cnt[f][0])
    cv2.line(img,start,end,[0,0,255],2)
    cv2.circle(img,far,5,(255,0,0),-1)


cv2.imshow("img",img)

cv2.waitKey(0)
cv2.destroyAllWindows()

    defects:
 [[[    0   102    46 29796]]

 [[  102   105   104   284]]

 [[  106   166   133 17825]]

 [[  167   173   170   225]]

 [[  173   181   180   262]]

 [[  181   189   188   208]]

 [[  189   248   238  5027]]

 [[  250   344   267  3672]]

 [[  345   347   346   114]]

 [[  347   428   390 20365]]

 [[  428   524   479 28489]]

 [[  524   526   525   186]]]

    cnt=contours[0]
#print(cnt) 随机图片会出现的问题，很小，捕捉不到凸包
#[[[256 366]]

# [[255 367]]]

    3. 几何学测试

    检测轮廓是否是凸形

    retval=cv2.isContourConvex(contour)

    retval: bool类型
    contour：需要判断的轮廓

    import cv2

img=cv2.imread("D:/part12/part12_8.bmp")

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)
#-----------------凸包

image1=img.copy()
hull=cv2.convexHull(contours[0])
cv2.polylines(image1,[hull],True,[0,255,0],2)
print("使用函数cv2.convexHull()构造的多边形是否是凸形的：",cv2.isContourConvex(hull))
cv2.imshow("result1",image1)

#-------------------逼近多边形
image2=img.copy()
epsilon=0.01*cv2.arcLength(contours[0],True)
approx=cv2.approxPolyDP(contours[0],epsilon,True)
image2=cv2.drawContours(image2,[approx],0,(0,0,255),2)
print("使用函数cv2.approxPolyDP()构造的多边形是否是凸形：",cv2.isContourConvex(approx))
cv2.imshow("result2",image2)

#-----------------------
cv2.waitKey(0)
cv2.destroyWindow()

    使用函数cv2.convexHull()构造的多边形是否是凸形的： True
使用函数cv2.approxPolyDP()构造的多边形是否是凸形： False

    点到轮廓的距离

    用来计算点到多边形(轮廓)的最短距离，也就是垂线距离

    retval=cv2.pointPolygonTest(contour,pt,measureDist)

    pt: 为待判定的点
    measureDist：True（距离，内正，上0，外负）；False（相对位置关系，1，-1，0）

    import cv2

img=cv2.imread("D:/part12/part12_6.bmp")

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

#---------------获取凸包
image1=img.copy()
hull=cv2.convexHull(contours[0])
cv2.polylines(image1,[hull],True,(0,255,0),2)
#print(hull)

#------------------内部点A到轮廓的距离
distA=cv2.pointPolygonTest(hull,(320,50),True)
font=cv2.FONT_HERSHEY_SIMPLEX
cv2.putText(image1,'A',(320,50),font,1,(0,255,0),3)
print("distA=",distA)

#-----------------轮廓上点B到轮廓的距离
distB=cv2.pointPolygonTest(hull,(432,100),True)
font=cv2.FONT_HERSHEY_SIMPLEX
cv2.putText(image1,'B',(432,100),font,1,(0,255,0),3)
print("distB=",distB)

#-----------------轮廓外点C到轮廓的距离
distC=cv2.pointPolygonTest(hull,(432,150),True)
font=cv2.FONT_HERSHEY_SIMPLEX
cv2.putText(image1,'C',(432,150),font,1,(0,255,0),3)
print("distC=",distC)

cv2.imshow("image1",image1)
cv2.waitKey(0)
cv2.destroyAllWindows()

    distA= 19.27341155003588
distB= -0.0
distC= -27.164164493555205

    点与轮廓的位置判断

    将参数改为False

    distA= 1.0
distB= 0.0
distC= -1.0

    4. 轮廓的特征值

    1.宽高比

    import cv2

img=cv2.imread("D:/part12/part12_1.bmp")

cv2.imshow("result",img)

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

x,y,w,h=cv2.boundingRect(contours[0])
cv2.rectangle(img,(x,y),(x+w,y+h),(255,255,255),3)
aspectRatio=float(w)/h
print(aspectRatio)
cv2.imshow("result",img)
cv2.waitKey(0)
cv2.destroyWindow()


    1.6503067484662577

    2. Extent

    轮廓面积(对象面积)/矩形边界面积

    import cv2

img=cv2.imread("D:/part12/part12_1.bmp")

cv2.imshow("result",img)

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

x,y,w,h=cv2.boundingRect(contours[0])
cv2.drawContours(img,contours[0],-1,(0,0,255),3)
cv2.rectangle(img,(x,y),(x+w,y+h),(255,0,0),3)
rectArea=w*h
cntArea=cv2.contourArea(contours[0])
extend=float(cntArea)/rectArea
print(extend)
cv2.imshow("result",img)
cv2.waitKey(0)
cv2.destroyWindow()

    0.6662371427919812

    3. Solidity

    轮廓面积/凸包面积

    import cv2

img=cv2.imread("D:/part12/part12_6.bmp")

cv2.imshow("result",img)

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

cv2.drawContours(img,contours[0],-1,(0,255,255),3)

hull=cv2.convexHull(contours[0])
cntArea=cv2.contourArea(contours[0])
hullArea=cv2.contourArea(hull)

cv2.polylines(img,[hull],True,(0,255,0),2)

solidity=float(cntArea)/hullArea

print(solidity)
cv2.imshow("result",img)
cv2.waitKey(0)
cv2.destroyWindow()

    0.7329527393863876

    4. 等效直径

    该值是与轮廓面积相等的圆形的直径

    import cv2
import numpy as np

img=cv2.imread("D:/part12/part12_1.bmp")

cv2.imshow("img",img)
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

cv2.drawContours(img,contours[0],-1,(0,0,255),3)
cntArea=cv2.contourArea(contours[0])
equiDiameter=np.sqrt(4*cntArea/np.pi)

print(equiDiameter)
cv2.circle(img,(100,100),int(equiDiameter/2),(0,0,255),3)
cv2.imshow("result",img)

cv2.waitKey(0)
cv2.destroyWindow()

    192.85878305271953

    5. 方向

    (x,y),(MA,ma),angle=cv2.fitEllipse(cnt)

    其中：
    (x,y)：椭圆的中心点
    (MA,ma) ：椭圆水平方向轴和垂直方向轴的长度
    angle：椭圆的旋转角度

    import cv2

img=cv2.imread("D:/part12/part12_1.bmp")

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

ellipse=cv2.fitEllipse(contours[0])
retval=cv2.fitEllipse(contours[0])

print("单个返回形式：")
print("retval=\n",retval)
print("三个返回形式：")
(x,y),(MA,ma),angle=cv2.fitEllipse(contours[0])

print("(x,y)=(",x,y,")")
print("(MA,ma)=(",MA,ma,")")
print("angle=",angle)

cv2.ellipse(img,ellipse,(0,0,255),2)
cv2.imshow("img",img)
cv2.waitKey(0)
cv2.destroyWindow()

    单个返回形式：
retval=
 ((246.05667114257812, 185.529541015625), (141.45005798339844, 282.24273681640625), 77.81553649902344)
三个返回形式：
(x,y)=( 246.05667114257812 185.529541015625 )
(MA,ma)=( 141.45005798339844 282.24273681640625 )
angle= 77.81553649902344

    6. 掩模和像素点

    6.1 使用Numpy函数获取轮廓像素点

    cv2.drawContours()轮廓宽度参数thickness设置为“-1”,即可获取特定对象的实心轮廓，即特定对象的掩模

    numpy.nonzero()
#返回数组内非零元素的位置，但其返回的是行，列分别显示的

    numpy.transpose()
#处理上述返回值，则得到这些点的(x,y)形式的坐标

    import numpy as np

a=np.zeros((5,5),dtype=np.uint8)

for time in range(10):
    i=np.random.randint(0,5)
    j=np.random.randint(0,5)
    a[i,j]=1

print("a=\n",a)
loc=np.transpose(np.nonzero(a))

print("a内非零位置",loc)

    a=
 [[1 0 0 0 0]
 [1 0 0 0 1]
 [0 1 0 0 1]
 [1 1 1 1 0]
 [0 0 0 0 0]]
a内非零位置 [[0 0]
 [1 0]
 [1 4]
 [2 1]
 [2 4]
 [3 0]
 [3 1]
 [3 2]
 [3 3]]

    使用numpy获取一个图像内的轮廓点位置

    import cv2
import numpy as np

img=cv2.imread("D:/part12/part12_1.bmp")

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

cnt=contours[0]

#-----------------------------------------绘制空心轮廓
mask1=np.zeros(gray.shape,np.uint8)
cv2.drawContours(mask1,[cnt],0,(255,255,255),2)
pixelpoints1=np.transpose(np.nonzero(mask1))
print("pixelpoints1.shape:",pixelpoints1.shape)
print("pixelpoints1:\n",pixelpoints1)
cv2.imshow("mask1",mask1)

#-----------------------------------------绘制实心轮廓
mask2=np.zeros(gray.shape,np.uint8)
cv2.drawContours(mask2,[cnt],0,(255,255,255),-1)
pixelpoints2=np.transpose(np.nonzero(mask2))
print("pixelpoints2.shape:",pixelpoints2.shape)
print("pixelpoints2:\n",pixelpoints2)
cv2.imshow("mask2",mask2)

cv2.waitKey(0)
cv2.destroyWindow()


cv2.waitKey(0)
cv2.destroyWindow()

    pixelpoints1.shape: (2708, 2)
pixelpoints1:
 [[111 276]
 [111 277]
 [111 278]
 ...
 [275 150]
 [275 151]
 [275 152]]
pixelpoints2.shape: (29571, 2)
pixelpoints2:
 [[112 277]
 [112 278]
 [112 279]
 ...
 [274 149]
 [274 150]
 [274 151]]
    6.1 使用Opencv函数获取轮廓像素点
    import cv2
import numpy as np

a=np.zeros((5,5),dtype=np.uint8)

for time in range(5):
    i=np.random.randint(0,5)
    j=np.random.randint(0,5)
    a[i,j]=1

print("a=\n",a)

loc=cv2.findNonZero(a)

print("a内非零值得位置：\n",loc)

    a=
 [[0 0 0 0 0]
 [0 0 1 0 0]
 [0 1 0 0 0]
 [0 0 0 0 0]
 [1 1 0 0 0]]
a内非零值得位置：
 [[[2 1]]

 [[1 2]]

 [[0 4]]

 [[1 4]]]

    import cv2
import numpy as np

img=cv2.imread("D:/part12/part12_1.bmp")

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)


cnt=contours[0]

#----------------绘制空心轮廓
mask1=np.zeros(gray.shape,dtype=np.uint8)
cv2.drawContours(mask1,[cnt],0,255,2)
pixelpoints1=cv2.findNonZero(mask1)
print("pixelpoints.shape",pixelpoints1.shape)
print("pixelpoints",pixelpoints1)
cv2.imshow("mask1",mask1)
#----------------绘制实心轮廓
mask2=np.zeros(gray.shape,dtype=np.uint8)
cv2.drawContours(mask2,[cnt],0,255,-1)
pixelpoints2=cv2.findNonZero(mask2)
print("pixelpoints.shape",pixelpoints2.shape)
print("pixelpoints",pixelpoints2)
cv2.imshow("mask2",mask2)

cv2.waitKey(0)
cv2.destroyWindow()

    最大值和最小值以及其位置

    minval,maxval,minloc,maxloc=cv2.minMaxLoc(imgray,mask)

    指定的对象内查找最大值，最小值以及其位置

    import cv2
import numpy as np

img=cv2.imread("D:/101.bmp")

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

cnt=contours[2]

mask=np.zeros(gray.shape,dtype=np.uint8)
mask=cv2.drawContours(mask,[cnt],-1,255,-1)

cv2.imshow("mask",mask)

minval,maxval,minloc,maxloc=cv2.minMaxLoc(gray,mask=mask)

print("minval:\n",minval)
print("maxval:\n",maxval)
print("minloc:\n",minloc)
print("maxloc:\n",maxloc)

#构造掩模获取感兴趣区域
masko=np.zeros(img.shape,dtype=np.uint8)
masko=cv2.drawContours(masko,[cnt],-1,(255,255,255),-1)
loc=cv2.bitwise_and(img,masko)

cv2.imshow("loc",loc)

cv2.waitKey(0)
cv2.destroyAllWindows()

    minval:
 127.0
maxval:
 255.0
minloc:
 (271, 148)
maxloc:
 (271, 147)

    6. 平均颜色及灰度

    mean_val=cv2.mean(in,mask=mask)

    import cv2
import numpy as np

img=cv2.imread("D:/101.bmp")

gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

cnt=contours[2]

mask=np.zeros(gray.shape,dtype=np.uint8)
cv2.drawContours(mask,[cnt],0,255,-1)
mean=cv2.mean(img,mask=mask)
print("mean=:",mean)
mask2=np.zeros(img.shape,dtype=np.uint8)
cv2.drawContours(mask2,[cnt],0,(255,255,255),-1)
loc=cv2.bitwise_and(img,mask2)

cv2.imshow("loc",loc)
cv2.waitKey(0)
cv2.destroyAllWindows()

    mean=: (133.34867032503166, 133.34867032503166, 133.34867032503166, 0.0)

    7. 极点

    获取某个对象内的极点，例如最左端，最右端，最上端，最下端

    leftmost=tuple(cnt[cnt[:,:,0].argmin()][0])
rightmost=tuple(cnt[cnt[:,:,0].argmax()][0])
topmost=tuple(cnt[cnt[:,:,1].argmax()][0])
bottommost=tuple(cnt[cnt[:,:,1].argmin()][0])

    import cv2
import numpy as np

img=cv2.imread("D:/part12/part12_6.bmp")
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,binary=cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
contours,hierarchy=cv2.findContours(binary,
                                    cv2.RETR_LIST,
                                    cv2.CHAIN_APPROX_SIMPLE)

cnt=contours[0]
mask=np.zeros(gray.shape,dtype=np.uint8)
cv2.drawContours(mask,[cnt],0,255,-1)
#---------------------------------计算极值
leftmost=tuple(cnt[cnt[:,:,0].argmin()][0])
rightmost=tuple(cnt[cnt[:,:,0].argmax()][0])
topmost=tuple(cnt[cnt[:,:,1].argmax()][0])
bottommost=tuple(cnt[cnt[:,:,1].argmin()][0])
#---------------------------------打印极值
print("leftmost:",leftmost)
print("rightmost:",rightmost)
print("topmost",topmost)
print("bottommost",bottommost)
#---------------------------------绘制说明文字
font=cv2.FONT_HERSHEY_SIMPLEX
cv2.putText(img,'A',leftmost,font,1,(0,0,255),2)
cv2.putText(img,'B',rightmost,font,1,(0,0,255),2)
cv2.putText(img,'C',topmost,font,1,(0,0,255),2)
cv2.putText(img,'D',bottommost,font,1,(0,0,255),2)

#----------------------------------绘制图像
cv2.imshow("img",img)

cv2.waitKey(0)
cv2.destroyAllWindows()

    leftmost: (104, 159)
rightmost: (434, 111)
topmost (126, 237)
bottommost (322, 27)

    相关文章

      网友评论

          本文标题：图像轮廓(2)

          本文链接：https://www.haomeiwen.com/subject/piayaltx.html

          延伸阅读

          深度阅读

          • 您也可以注册成为美文阅读网的作者，发表您的原创作品、分享您的心情！

          栏目导航

          热点阅读

          关于我们|服务条款|联系我们|图像轮廓(2)|投稿指南|网站地图|RSS订阅|排版工具|手机版

          提供经典美文摘抄,优美散文欣赏,现代诗歌精选,短篇小说,心情随笔,表白情书范文,故事会在线阅读欣赏

          Copyright © 2014-2023 Haomeiwen.com All Rights Reserved. 好美文阅读网 版权所有

          备案信息：桂公网安备 45052102000051号 · 桂ICP备13007215号-3

          本站所收录作品、热点评论等信息部分来源互联网，目的只是为了系统归纳学习和传递资讯

          所有作品版权归原创作者所有，与本站立场无关，如不慎侵犯了你的权益，请联系我们告知，我们将做删除处理！