如何找到像素白色斑点的像素位置

Question

I have managed to detect a blue square from this image :

The mask has just black and white. I want to know the position of the white block, ie its middle point.

My question is : How do I detect the middle point of the blue square in the picture ?

I got the following code from the internet :

# import the necessary packages
import numpy as np
import cv2
def detectColouredObject(FILENAME):
    # load the image
    image = cv2.imread(FILENAME)

    # THE COLOURS ARE IN RGB
    lower_blue = np.array([50, 0, 0])
    upper_blue = np.array([255, 50, 50])

    # loop over the boundaries
    #    for (lower, upper) in boundaries:
        # create NumPy arrays from the boundaries
    lower = np.array(lower_blue, dtype = "uint8")
    upper = np.array(upper_blue, dtype = "uint8")

    # find the colors within the specified boundaries and apply
    # the mask
    mask = cv2.inRange(image, lower, upper)
    maskWidth, maskHeight = mask.shape[:2]

    cv2.imshow("mask ", mask)
    npImg = np.asarray( mask )  # No copying takes place

    coordList = np.argwhere( npImg == 255 )
    cv2.imshow("mask1 ", coordList)
    print coordList
    xmin = np.amin(coordList,axis=0)
    xmax = np.amax(coordList,axis=0)
    ymax = np.amax(coordList,axis=1)
    xStart = xmin[0]
    xEnd = xmax[0]

    output = cv2.bitwise_and(image, image, mask = mask)
    width, height = output.shape[:2]
    midpoint = width / 2


    # show the images
    cv2.imshow("images", np.hstack([image, output]))
    cv2.waitKey(0)

Thank you to your help

Answer 1

You're on to the right idea by thresholding and coming up with a nice white blob, the next step is to use contours and then image moment analysis .

The system treats pixels as having 'mass' - ie white is heavier than black.

Fun fact: It's actually a direct parallel to the mechanical process of finding a (planar) centre of mass in a solid - but discretised over pixels (ie summation, not integration)

Answer 2

上面的答案很好并且完全正确，但是为简单起见，您可能会很满意： imerode ，只要图片上有白色，就可以使用。