opencv跟踪图像中的点

Question

我试图在视网膜移动时跟踪视网膜图像上的点。目前，我正在使用OpenCV的模板匹配方法来识别点周围的区域，然后在下一幅图像中定位该区域，以识别点所在的位置

我有两张视网膜图像：

在第一幅图像中，我任意选择了3个我想要检测的点：

然后，我使用以下代码跟踪未来图像中的点：

s = 50
folder = 'P2/'
num_images = 2
#desired tracking point: (x, y)
x1, y1 = 157 - s//2, 130 - s//2
x2, y2 = 182 - s//2, 59 - s//2
x3, y3 = 221 - s//2, 125 - s//2
#template is a 100x100 square with (x,y) as top left corner
template = cv2.imread(folder + '1.png',0)
templates = [template[y1:y1+s, x1:x1+s], template[y3:y3+s, x3:x3+s], template[y2:y2+s, x2:x2+s]]
w, h = templates[0].shape[::-1]

# All the 6 methods for comparison in a list
methods = ['cv2.TM_CCOEFF', 'cv2.TM_CCOEFF_NORMED', 'cv2.TM_CCORR',
            'cv2.TM_CCORR_NORMED', 'cv2.TM_SQDIFF', 'cv2.TM_SQDIFF_NORMED']

for i in range(1, num_images+1):
    img = cv2.imread(folder + str(i)+'.png',0)
    meth = 'cv2.TM_CCOEFF_NORMED' #works the best
    method = eval(meth)

    x_pts = []
    y_pts = []
    for tem in templates:
        # Apply template Matching
        res = cv2.matchTemplate(img,tem,method)
        min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)

        # If the method is TM_SQDIFF or TM_SQDIFF_NORMED, take minimum
        if method in [cv2.TM_SQDIFF, cv2.TM_SQDIFF_NORMED]:
            top_left = min_loc
        else:
            top_left = max_loc
        bottom_right = (top_left[0] + w, top_left[1] + h)

        #cv2.rectangle(img,top_left, bottom_right, 255, 2)
        x = top_left[0] + s//2
        y = top_left[1] + s//2
        x_pts += [x]
        y_pts += [y]
        #cv2.rectangle(img,(x, y),(x+1,y+1),(0,255,0),2) #new (x, y) point
        print(top_left)
    plt.plot(x_pts, y_pts,'ro-')
    plt.imshow(img, cmap='gray')
    plt.title(calculate(x_pts,y_pts)), plt.xticks([]), plt.yticks([])
    plt.suptitle(meth)

    plt.show()

现在，下图中的点如下所示：

如果点周围的区域是唯一的，这种跟踪方法是准确的，我想知道是否有更好的方法跟踪图像中的点？例如，如果我正在跟踪图像中较难分辨的模糊区域