Btukfyl

Bascially, I have the mouse_match opencv code that takes a directory path as an argument, loops over the images in the directory, then lets the user select -by mouse- a portion of an image, then performs template matching against the full image.
It has the following undesirable functionalities:
1. Takes input as command line arguments
2. Returns a black image with bright spots marking the points where the highest similarity was found.

Is it possible to create a GUI using opencv methods and classes so that it achieves:
1. the user should be able to do multiple selections, from one image or from other images in the directory also.
2. template matching for all objects in all images.
3. return the directory images with bounding boxes enclosing the occurrences of selected objects.
4. show some statistics about the occurrences, ex: each object against number of occurrences.

Any help will be appreciated :)
Even pesudocode-like instructions will be handy..

This is the code that I have in hand:

"""

mouse_and_match.py [-i path | --input path: default ./]

Demonstrate using a mouse to interact with an image:

 Read in the images in a directory one by one

 Allow the user to select parts of an image with a mouse

 When they let go of the mouse, it correlates (using matchTemplate) that patch with the image.

 ESC to exit

"""



import numpy as np

from math import *

import sys

import os

import glob

import argparse

import cv2 as cv



drag_start = None

sel = (0,0,0,0)



def onmouse(event, x, y, flags, param):

    global drag_start, sel

    if event == cv.EVENT_LBUTTONDOWN:

        drag_start = x, y

        sel = 0,0,0,0

    elif event == cv.EVENT_LBUTTONUP:

        if sel[2] > sel[0] and sel[3] > sel[1]:

            patch = gray[sel[1]:sel[3],sel[0]:sel[2]]

            result = cv.matchTemplate(gray,patch,cv.TM_CCOEFF_NORMED)

            result = np.abs(result)**3

            val, result = cv.threshold(result, 0.01, 0, cv.THRESH_TOZERO)

            result8 = cv.normalize(result,None,0,255,cv.NORM_MINMAX,cv.CV_8U)

            cv.imshow("result", result8)

        drag_start = None

    elif drag_start:

        #print flags

        if flags & cv.EVENT_FLAG_LBUTTON:

            minpos = min(drag_start[0], x), min(drag_start[1], y)

            maxpos = max(drag_start[0], x), max(drag_start[1], y)

            sel = minpos[0], minpos[1], maxpos[0], maxpos[1]

            img = cv.cvtColor(gray, cv.COLOR_GRAY2BGR)

            cv.rectangle(img, (sel[0], sel[1]), (sel[2], sel[3]), (0,255,255), 1)

            cv.imshow("gray", img)

        else:

            print("selection is complete")

            drag_start = None



if __name__ == '__main__':

    parser = argparse.ArgumentParser(description='Demonstrate mouse interaction with images')

    parser.add_argument("-i","--input", default='./', help="Input directory.")

    args = parser.parse_args()

    path = args.input



    #cv.namedWindow("gray",1)

    cv.namedWindow("gray", cv.WINDOW_AUTOSIZE)

    cv.setMouseCallback("gray", onmouse)

    '''Loop through all the images in the directory'''

    for infile in glob.glob( os.path.join(path, '*.*') ):

        ext = os.path.splitext(infile)[1][1:] #get the filename extenstion

        if ext == "png" or ext == "jpg" or ext == "bmp" or ext == "tiff" or ext == "pbm":

            print(infile)



            img=cv.imread(infile,1)

            if img.all() == None:

                continue

            sel = (0,0,0,0)

            drag_start = None

            gray1=cv.cvtColor(img, cv.COLOR_BGR2GRAY)

            ret,gray=cv.threshold(gray1,127,255,cv.THRESH_BINARY)

            cv.imshow("gray",gray)

            if (cv.waitKey() & 255) == 27:

                break

    cv.destroyAllWindows()  

Here is a program that integrates openCV with a GUI. It identifies and labels objects using YOLO.

You will need to install PySimpleGUI for the GUI portion. It utilizes tkinter which is built into Python. To keep things really simple, PySimpleGUI is a single .py file should you not want to do pip installs.

Video of the program running.