img2wav.py

import os
import sys
import cv2
import numpy as np
from scipy.io import wavfile


def image_to_audio(image_path, output_wav):
    img = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
    height, width = img.shape
    f = np.fft.fft2(img)
    fshift = np.fft.fftshift(f)
    inverse_f = np.fft.ifftshift(fshift)
    img_reconstructed = np.fft.ifft2(inverse_f)
    audio_signal = np.real(img_reconstructed).flatten()
    audio_signal_normalized = np.int16(
        (audio_signal / np.max(np.abs(audio_signal))) * 32767
    )
    sample_rate = 44100
    wavfile.write(output_wav, sample_rate, audio_signal_normalized)
    print(f"Audio saved as {output_wav}")


if __name__ == "__main__":
    if len(sys.argv) < 2:
        print("Usage: python script_name.py image_file")
        sys.exit(1)

    img_file = sys.argv[1]
    basename = os.path.basename(img_file)

    wav_file = basename + ".wav"

    image_to_audio(img_file, wav_file)