如何在opencv中计算黑色像素轮廓上的固体度

Question

坚固性按面积/凸壳面积之比计算:

#calculating area from contour
area = cv2.contourArea(unicocnt)
#calculating hull and hull area
hull = cv2.convexHull(unicocnt) 
hull_area = cv2.contourArea(hull)
#solidity
solidity = float(area)/hull_area

对圆形图像进行计算，我得到的所有值都接近于1，所以我认为，当我计算轮廓面积时，我计算的是圆圈内的面积，而不考虑圆圈内是否为白色像素(轮廓为黑色)

图片示例:https://docs.google.com/file/d/0ByS6Z5WRz-h2b0JITFB4aHR0OWc/edit?usp=sharing

代码:

nomeimg = 'Riscalate2/JPEG/e (5).jpg'
img = cv2.imread(nomeimg)
gray = cv2.imread(nomeimg,0)#convert grayscale adn binarize
element = cv2.getStructuringElement(cv2.MORPH_CROSS,(6,6)) 
graydilate = cv2.erode(gray, element) #imgbnbin
cv2.imshow('image',graydilate)
cv2.waitKey(0)
ret,thresh = cv2.threshold(graydilate,127,255,cv2.THRESH_BINARY_INV)   # binarize
imgbnbin = thresh
cv2.imshow('bn',thresh)
cv2.waitKey()
#element = cv2.getStructuringElement(cv2.MORPH_CROSS,(2,2))
#element = np.ones((11,11),'uint8')

contours, hierarchy = cv2.findContours(imgbnbin, cv2.RETR_TREE ,cv2.CHAIN_APPROX_SIMPLE)
print(len(contours))

# Take only biggest contour basing on area
Areacontours = list()
calcarea = 0.0
unicocnt = 0.0
for i in range (0, len(contours)):
    area = cv2.contourArea(contours[i])
    #print("area")
    #print(area)
    if (area > 90 ):  #con 90 trova i segni e togli puntini
        if (calcarea<area):
            calcarea = area
            unicocnt = contours[i]
#calculating area from contour
area = cv2.contourArea(unicocnt)
#calculating hull and hull area
hull = cv2.convexHull(unicocnt) 
hull_area = cv2.contourArea(hull)
#solidity
solidity = float(area)/hull_area

我是这样做的:

ColoredArea = 0
for i in range(0,len(imgbnbin)):
    a = imgbnbin[i]
    for j in range (0, len(a)):
        if (cv2.pointPolygonTest(hull, unicocnt) >= 0):        
            if (getPixel(x,y) == black):
                ColoredArea = ColoredArea +1; 

出现错误:

if (cv2.pointPolygonTest(hull, unicocnt) >= 0):
TypeError: Required argument 'measureDist' (pos 3) not found

Answer 1

试试下面的算法(伪代码):

int ColoredArea = 0;
for (x=MinX to MaxX)
   for (y=MinY to MaxY)
        if (cv2.pointPolygonTest(hull, pt(x,y)) >= 0)
           if (getPixel(x,y) == black)
               ColoredArea++;