Extracts vallejo colors from their released color charts and puts into Real Color Mixer JSON format

vallejo_color_extract.py

import cv2
import pytesseract
from pytesseract import Output
import numpy as np
import re
import json

# Create a JSON object with a JSON array
palette = {
    "name": "Vallejo Panzer Aces",
    "colors": [
    ],
    "colorsToMix": []
}

# Load the image
# image = cv2.imread('CC070-Model_Color-Rev18-baja-2.png')
# image = cv2.imread('CC064-MechaColor-Rev03-baja-1.png')
# image = cv2.imread('CC072-Game_Color-Rev20-baja-1.png')
# image = cv2.imread('CC082-Game_Air-Rev05-baja-2.png')
image = cv2.imread('CC077-Panzer_Aces-Rev12-baja-1.png')

# Convert the image to grayscale
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

# Apply adaptive thresholding to the grayscale image
thresh = cv2.adaptiveThreshold(
    gray, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY_INV, 11, 2)

# Find contours in the thresholded image
contours, _ = cv2.findContours(
    thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

# Set pytesseract configuration
custom_config = r'--oem 3 --psm 6'

# Iterate through the contours
count = 0
named = 0
for contour in contours:
    # Get the bounding rectangle for each contour
    x, y, w, h = cv2.boundingRect(contour)

    # Check if the contour is likely to be a rectangle based on its aspect ratio
    aspect_ratio = float(w) / h
    if w > 80 and h > 10:
        # Extract the color region
        color_region = image[y + 5:y + h - 10, x + 5:x + w - 10]

        # Calculate the average BGR values of the color inside the rectangle
        avg_color_bgr = np.mean(
            np.mean(color_region, axis=0), axis=0).astype(int)

        # Convert the average color to RGB format
        avg_color_rgb = np.flip(avg_color_bgr).tolist()
        print(f"Average RGB values: {avg_color_rgb}")

        # Extract the text region
        text_region = gray[y + h + 3:y + h + int(1 * h) + 3, x:x + w]
        # Recognize the text using pytesseract
        text = pytesseract.image_to_string(
            text_region, config=custom_config).strip()

        text_lines = text.splitlines()

        while ("" in text_lines):
            text_lines.remove("")

        if (len(text_lines) > 1):
            # Print the recognized text
            color_name = re.sub(r'[^a-zA-Z0-9 ]', '', text_lines[1])
            print(f"Color name: {color_name}")

            # Draw the rectangle on the original image
            cv2.rectangle(image, (x, y), (x + w, y + h), (0, 255, 0), 2)
            count = count + 1
            if len(text_lines[1].strip()) == 0:
                cv2.rectangle(image, (x, y + h + 3), (x+w, y +
                              h + 3 + int(1 * h)), (0, 0, 255), 2)
                print(f"Found lines: {text_lines}")
                input("Press Enter to continue...")
            else:
                cv2.rectangle(image, (x, y + h + 3), (x+w, y +
                              h + 3 + int(1 * h)), (255, 0, 0), 2)
                named = named + 1

                new_color = {
                    "r": avg_color_rgb[0],
                    "g": avg_color_rgb[1],
                    "b": avg_color_rgb[2],
                    "group": "basic",
                    "name": text_lines[1].strip()
                }

                palette["colors"].append(new_color)
cv2.imwrite("rectangles.png", image)
print(f"Found {count}, Named {named}")
cv2.waitKey(0)
cv2.destroyAllWindows()

with open('vallejo_panzer_aces.json', 'w') as file:
    json.dump(palette, file, indent=4)
print("Done!")