Is it possible to make my Python code faster?

This is my code:

img_gray = loadimagein_gray(im1)
count = 0
for X in path:
    template = cv2.imread(X)
    found = False
    w, h = template.shape[:-1]
    res = cv2.matchTemplate(img_gray, template, cv2.TM_CCOEFF_NORMED)
    threshold =0.85
    loc = np.where(res >= threshold)
    count = count + 1
    print(count)
    if found == False:
        for pt in zip(loc[::-1]): # Switch collumns and rows cv2.rectangle(img_gray, pt, (pt[0] + w, pt[1] + h), (0, 0, 255) 

440/5000

Compare with this code whether an image in this case img_gray has a Watermark template or X. I have about 3000 watermark logos and one website https://buyessayclub.com/ with custom design and this whole code takes about 10 to a high probability can be found that no watermark in the image is present. Since I wanted to take 10 minutes too long I ask there is a possibility to make the code faster. Here is the highest improvement potential for this line:

res = cv2.matchTemplate(img_gray, template, cv2.TM_CCOEFF_NORMED)