OpenCV task of detect object in image

Question

I'm trying to detect the fragment in image, represented in this MATHLAB example.

And I use OpenCV-library.

  import cv2
  import numpy as np
  from imutils.object_detection import non_max_suppression

  # Reading the image and the template
  img = cv2.imread('SourceImage.png')
  temp = cv2.imread('TargetFragment.png')
  # save the image dimensions
  W, H = temp.shape[:2]
  # Define a minimum threshold
  thresh = 0.4
  # Converting them to grayscale
  img_gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
  temp_gray = cv2.cvtColor(temp, cv2.COLOR_BGR2GRAY)
  # Passing the image to matchTemplate method
  match = cv2.matchTemplate(
            image=img_gray, templ=temp_gray,
            method=cv2.TM_CCOEFF_NORMED)
  # Select rectangles with
  # confidence greater than threshold
  (y_points, x_points) = np.where(match >= thresh)
  # initialize our list of rectangles
   boxes = list()
  # loop over the starting (x, y)-coordinates again
  for (x, y) in zip(x_points, y_points):
             # update our list of rectangles
             boxes.append((x, y, x + W, y + H))
  # apply non-maxima suppression to the rectangles
  # this will create a single bounding box
  boxes = non_max_suppression(np.array(boxes))
  # loop over the final bounding boxes
  for (x1, y1, x2, y2) in boxes:
         # draw the bounding box on the image
         cv2.rectangle(img, (x1, y1), (x2, y2),
                   (255, 0, 0), 3)
  cv2.imwrite('result.png', img)

Big Image is:

Target fragment to detect is:

But 2 areas are detected, instead of one. One of this areas doesn't contain the target fragment at all:

What did I miss?

Answer 1

This is not a great task for matchTemplate. matchTemplate is better suited for tasks where an almost exact copy of the template exists in the target image.

Checkout this answer for an example of feature matching: How to match and align two images using SURF features (Python OpenCV )?

Good luck!