scaled_template_matching_multi.py

"""
Usage: python scaled_template_matching_multi.py --template tmp.jpg --image img.jpg

Installation: 
pip install opencv-contrib-python numpy imutils

https://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_imgproc/py_template_matching/py_template_matching.html#template-matching-with-multiple-objects
https://stackoverflow.com/questions/49464639/opencv-matchtemplate-threshold-values-for-different-methods
"""
import glob
import argparse

import numpy as np
import imutils
import cv2

# construct the argument parser and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument(
    "-t", "--template", required=True, help="Path to template image")
# ap.add_argument(
#     "-i", "--images", required=True,
#     help="Path to images where template will be matched")
ap.add_argument(
    "-i", "--image", required=True,
    help="Path to the image where template will be matched")
ap.add_argument(
    "-v", "--visualize",
    help="Flag indicating whether or not to visualize each iteration")
args = vars(ap.parse_args())

tmp_path = args['template']
img_path = args['image']
res_path = '{}_{}.jpg'.format(
    tmp_path.strip('.jpg').strip('.jpeg').strip('.png'),
    img_path.split('/')[-1].strip('.jpg').strip('.jpeg').strip('.png'),
)

# load the image image, convert it to grayscale, and detect edges
template = cv2.imread(args["template"])
template = imutils.resize(template, width=200)
template = cv2.cvtColor(template, cv2.COLOR_BGR2GRAY)
template = cv2.Canny(template, 50, 200)
(tH, tW) = template.shape[:2]

# loop over the images to find the template in
# for imagePath in glob.glob(args["images"] + "/*.jpg"):
for imagePath in glob.glob(args["image"]):
    # load the image, convert it to grayscale, and initialize the
    # bookkeeping variable to keep track of the matched region
    image = cv2.imread(imagePath)
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    match_list = []
    # loop over the scales of the image
    for scale in np.linspace(1, 2, 10)[::-1]:
        # resize the image according to the scale, and keep track
        # of the ratio of the resizing
        resized = imutils.resize(gray, width=int(gray.shape[1] * scale))
        r = gray.shape[1] / float(resized.shape[1])
        # if the resized image is smaller than the template, then break
        # from the loop
        if resized.shape[0] < tH or resized.shape[1] < tW:
            break

        edged = cv2.Canny(resized, 50, 200)
        res = cv2.matchTemplate(edged, template, cv2.TM_CCOEFF_NORMED)
        # https://docs.opencv.org/3.4.0/df/dfb/group__imgproc__object.html#ga3a7850640f1fe1f58fe91a2d7583695d
        threshold = 0.094
        loc = np.where(res >= threshold)
        # print(loc)
        for pt in zip(*loc[::-1]):
            match_list.append((pt, r))

    print(len(match_list))
    print(match_list)
    # TODO: group similar points and averge them

    # unpack the bookkeeping variable and compute the (x, y) coordinates
    # of the bounding box based on the resized ratio
    for found in match_list:
        pt, r = found
        (startX, startY) = (int(pt[0] * r), int(pt[1] * r))
        (endX, endY) = (int((pt[0] + tW) * r), int((pt[1] + tH) * r))
        # draw a bounding box around the detected result and display the image
        cv2.rectangle(image, (startX, startY), (endX, endY), (0, 0, 255), 2)

    cv2.imshow("Image", image)
    cv2.waitKey(0)
    cv2.destroyAllWindows()

    print(f'saving res img to {res_path}')
    cv2.imwrite(res_path, image)