Matching a pattern between an image and template using Python OpenCV and SIFT

Question

I try to match a pattern of seven dots similar to letter H on template image with similar pattern which could be found on actual image. Currently SIFT is correctly recognising 2 of 7 dots and after many tries I ran out of ideas.

Input image

pattern to match

import cv2
import numpy as np
import matplotlib.pyplot as plt

image = cv2.imread('plaza.jpg')
image1 = cv2.imread('wzor.jpg')

blur = cv2.GaussianBlur(image, (9, 9), 0)
blur1 = cv2.GaussianBlur(image1, (9, 9), 0)
gray = cv2.cvtColor(blur, cv2.COLOR_BGR2GRAY)
gray1 = cv2.cvtColor(blur1, cv2.COLOR_BGR2GRAY)

edged = cv2.Canny(gray, 30, 200)
edged1 = cv2.Canny(gray1, 30, 200)
edged2=edged
contours, hierarchy = cv2.findContours(edged, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)
contours1, hierarchy1 = cv2.findContours(edged1, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)


cont = cv2.drawContours(image, contours, -1, (0, 255, 0), 3)
cont1 = cv2.drawContours(image1, contours1, -1, (0, 255, 0), 3)


sift = cv2.SIFT_create()


kp1, des1 = sift.detectAndCompute(edged,None)
kp2, des2 = sift.detectAndCompute(edged1,None)

bf = cv2.BFMatcher()
matches = bf.knnMatch(des1,des2, k=2)


good = []
for m,n in matches:
   if m.distance < 0.7*n.distance:
      good.append([m])

img3 = cv2.drawMatchesKnn(edged, kp1, edged1, kp2, good, None, flags=2)
plt.imshow(img3), plt.show()
img4 = cv2.drawMatchesKnn(image, kp1, image1, kp2, good, None, flags=2)
plt.imshow(img4), plt.show()

Outcome image

Outcome image 2

I want to achieve proper matching of all 7 points. Points size and placement is not the same on template and actual image. I want to find similar patterns but not only exactly the same ones. In future I want to try to apply this in live-time footage.