Restore Old Photos with AI: Colorize + Enhance Faces via API

Old family photos fade and lose detail over time. Restoring them used to require hours of Photoshop work. With two AI APIs, you can automate the entire process: colorize a black and white photo, then enhance the faces.

The Pipeline

1. Colorize: send the B&W photo to the Image Colorization API
2. Enhance faces: pass the colorized result to the Face Restoration API
3. Download: save the final restored image

The second API call uses the output URL from the first step directly. No need to download and re-upload.

Python Code

import requests
from pathlib import Path

API_KEY = "YOUR_API_KEY"
COLORIZE_URL = "https://photocolorizer-ai.p.rapidapi.com/colorize-photo"
RESTORE_URL = "https://face-restoration.p.rapidapi.com/enhance-face"


def restore_old_photo(image_path, output_path="restored.jpg"):
    # Step 1: Colorize
    with open(image_path, "rb") as f:
        color_resp = requests.post(
            COLORIZE_URL,
            headers={
                "x-rapidapi-key": API_KEY,
                "x-rapidapi-host": "photocolorizer-ai.p.rapidapi.com",
            },
            files={"image": f},
        )
    colorized_url = color_resp.json()["image_url"]

    # Step 2: Enhance faces
    face_resp = requests.post(
        RESTORE_URL,
        headers={
            "x-rapidapi-key": API_KEY,
            "x-rapidapi-host": "face-restoration.p.rapidapi.com",
            "Content-Type": "application/x-www-form-urlencoded",
        },
        data={"image_url": colorized_url, "mode": "all"},
    )
    restored = face_resp.json()

    # Step 3: Download
    img_data = requests.get(restored["image_url"]).content
    Path(output_path).write_bytes(img_data)
    return restored


result = restore_old_photo("grandma_1950.jpg")
print(f"Restored {result['faces_detected']} face(s)")

Why Two APIs?

Colorization and face restoration solve different problems. Colorization models predict what colors belong in a grayscale image. Face restoration models reconstruct facial features from low-quality inputs. Neither does the other's job. By chaining both, you get color and sharp faces in one pipeline.

Batch Processing

from concurrent.futures import ThreadPoolExecutor, as_completed
import os

def process_folder(input_dir, output_dir, max_workers=4):
    os.makedirs(output_dir, exist_ok=True)
    photos = [f for f in os.listdir(input_dir) if f.lower().endswith((".jpg", ".png"))]

    with ThreadPoolExecutor(max_workers=max_workers) as pool:
        futures = {
            pool.submit(restore_old_photo, f"{input_dir}/{p}", f"{output_dir}/restored_{p}"): p
            for p in photos
        }
        for fut in as_completed(futures):
            print(f"Done: {futures[fut]}")

process_folder("old_photos/", "restored/")

Use Cases

• Genealogy apps: users upload old family photos, get colorized and enhanced versions
• Museum digitization: batch process historical photo archives
• Photo restoration SaaS: wrap the pipeline behind a web interface

👉 Read the full tutorial with before/after results and API vs GFPGAN comparison