I tried moving the image to the specified folder by right-clicking and left-clicking

Introduction

I made it while thinking that I should be able to classify images easily.

What I made

I made a program to move images to a specified folder by right-clicking and left-clicking.

The image was downloaded from the standard image database SIDBA of Kanagawa Institute of Technology.

When you run the program, 5 images will be displayed.

I clicked on the image of the earth. The window will be updated.

The image of the clicked earth has been moved. Since it was a left click this time, it is output to the L folder.

code

imageMover.py

import cv2
import glob
import shutil
import os
import numpy as np

#Image size is unified at 200px
size = (200, 200)  

name = []   #file name
data = []   #File data
coordinates = []    #Coordinates when clicked

#Create a destination folder. L and R correspond to left click and right click respectively.
os.makedirs('./L', exist_ok=True)
os.makedirs('./R', exist_ok=True)

#Read file
for file in glob.glob('*.bmp'):
    img = cv2.imread(file)
    img = cv2.resize(img, size) 
    name.append(file) 
    data.append(img)

#Function executed when mouse click
def click_event(event, x, y, flags, param):
    if event == cv2.EVENT_LBUTTONDOWN:
        coordinates[0:3] = [x, y, 'L']
    if event == cv2.EVENT_RBUTTONDOWN:
        coordinates[0:3] = [x, y, 'R']

#Number of images
hasWindow = len(data)

#Execute when the number of remaining images is 1 or more
while hasWindow > 0:
    img = cv2.hconcat(data[:5])

    while(1):
        cv2.imshow('img', img)
        cv2.setMouseCallback("img", click_event) #When clicked
        
        #Get the coordinates when clicked and move the corresponding image
        if len(coordinates) != 0:
            n = coordinates[0]//200
            shutil.move(name[n], coordinates[2]+'/'+name[n])
            print(F'folder{coordinates[2]}To{name[n]}Moved')
            data.pop(n)
            name.pop(n)
            coordinates = []
            hasWindow -= 1 #Update the remaining number of images
            break
        
        #Initialize coordinates
        coordinates = []    

        #Is there a key input
        key = cv2.waitKey(100) & 0xff

        #Close window when keyboard or x is pressed
        if key != 255 or cv2.getWindowProperty('img', cv2.WND_PROP_AUTOSIZE) == -1:
            cv2.destroyAllWindows()
            exit()

I have also uploaded it to Github.

Conclusion

Isn't manual faster?

in conclusion

It seems that you can devise more by increasing the number of images displayed.

Thank you for watching until the end. We look forward to your suggestions and comments.