import cv2
import pytesseract
import re
import sys
import os

pytesseract.pytesseract.tesseract_cmd = "/usr/bin/tesseract"

def detect_plate_number(image_path):
    image = cv2.imread(image_path)
    if image is None:
        print(f"Fehler: Datei '{image_path}' konnte nicht geladen werden.")
        return None

    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    blurred = cv2.GaussianBlur(gray, (5, 5), 0)

    # Kanten finden
    edges = cv2.Canny(blurred, 100, 200)
    contours, _ = cv2.findContours(edges.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

    contours = sorted(contours, key=cv2.contourArea, reverse=True)
    plate_contour = None

    for contour in contours:
        epsilon = 0.02 * cv2.arcLength(contour, True)
        approx = cv2.approxPolyDP(contour, epsilon, True)

        if len(approx) == 4:  # Rechteck gefunden
            plate_contour = approx
            break

    if plate_contour is not None:
        x, y, w, h = cv2.boundingRect(plate_contour)
        plate_image = gray[y:y + h, x:x + w]

        # --- Linken Rand (EU-Bereich) abschneiden ---
        cut_x = int(w * 0.15)
        plate_image = plate_image[:, cut_x:]

        # --- Vergrößern ---
        plate_image = cv2.resize(plate_image, None, fx=2, fy=2, interpolation=cv2.INTER_CUBIC)

        # --- Threshold-Varianten ---
        _, thresh_otsu = cv2.threshold(plate_image, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
        thresh_adapt = cv2.adaptiveThreshold(
            plate_image, 255,
            cv2.ADAPTIVE_THRESH_GAUSSIAN_C,
            cv2.THRESH_BINARY,
            11, 2
        )

        # --- Zwischenschritte speichern ---
        base_name = os.path.splitext(os.path.basename(image_path))[0]
        cv2.imwrite(f"{base_name}_plate_otsu.png", thresh_otsu)
        cv2.imwrite(f"{base_name}_plate_adapt.png", thresh_adapt)
        print(f"Zwischenschritte gespeichert: {base_name}_plate_otsu.png, {base_name}_plate_adapt.png")

        # --- OCR Settings ---
        configs = {
            "otsu_psm7": ("--psm 7 -c tessedit_char_whitelist=ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789- ", thresh_otsu),
            "otsu_psm8": ("--psm 8 -c tessedit_char_whitelist=ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789- ", thresh_otsu),
            "adapt_psm7": ("--psm 7 -c tessedit_char_whitelist=ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789- ", thresh_adapt),
            "adapt_psm8": ("--psm 8 -c tessedit_char_whitelist=ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789- ", thresh_adapt),
        }

        results = {}
        for label, (config, img) in configs.items():
            txt = pytesseract.image_to_string(img, config=config).strip()
            results[label] = txt
            print(f"OCR Ergebnis [{label}]: {txt}")

        # bestes Ergebnis wählen (längster String)
        plate_number = max(results.values(), key=len) if any(results.values()) else None

        if plate_number:
            plate_number_clean = plate_number.replace(" ", "").replace("-", "")

            match_de = re.match(r"^([A-Z]{1,3})([A-Z]{1,2})([0-9]{1,4})$", plate_number_clean)
            match_nl = re.match(r"^([A-Z]{1,2})([0-9]{2,3})([A-Z]{1,2})$", plate_number_clean)

            if match_de:
                plate_number = f"{match_de.group(1)} {match_de.group(2)} {match_de.group(3)}"
            elif match_nl:
                plate_number = f"{match_nl.group(1)} {match_nl.group(2)} {match_nl.group(3)}"

            return plate_number.strip()

    return None


if __name__ == "__main__":
    if len(sys.argv) < 2:
        print("Aufruf: python3 license.py <bilddatei>")
        sys.exit(1)

    image_path = sys.argv[1]
    plate_number = detect_plate_number(image_path)
    print("Detected Plate Number:", plate_number)