| | import os
|
| | import sys
|
| | import argparse
|
| | import json
|
| | from datetime import datetime
|
| | import requests
|
| | from io import BytesIO
|
| |
|
| | import numpy as np
|
| | from PIL import Image
|
| | from skimage.color import rgb2lab, deltaE_ciede2000
|
| | from basicsr.metrics.niqe import calculate_niqe
|
| |
|
| | def download_image(path_or_url: str) -> Image.Image:
|
| | """Download or read local image, return PIL.Image in RGB mode."""
|
| | if path_or_url.startswith(('http://', 'https://')):
|
| | resp = requests.get(path_or_url, timeout=10)
|
| | if resp.status_code != 200:
|
| | raise ValueError(f"Failed to download image: {path_or_url} (status code {resp.status_code})")
|
| | data = BytesIO(resp.content)
|
| | else:
|
| | if not os.path.isfile(path_or_url):
|
| | raise ValueError(f"File does not exist: {path_or_url}")
|
| | data = path_or_url
|
| | try:
|
| | return Image.open(data).convert('RGB')
|
| | except Exception as e:
|
| | raise ValueError(f"Failed to open image: {e}")
|
| |
|
| | def compute_ciede2000(ref_img: Image.Image, test_img: Image.Image) -> float:
|
| | arr_ref = np.asarray(ref_img, dtype=np.float32) / 255.0
|
| | arr_test = np.asarray(test_img, dtype=np.float32) / 255.0
|
| | lab_ref = rgb2lab(arr_ref)
|
| | lab_test = rgb2lab(arr_test)
|
| | delta = deltaE_ciede2000(lab_ref, lab_test)
|
| | return float(np.mean(delta))
|
| |
|
| | def compute_niqe(img: Image.Image) -> float:
|
| | arr = np.asarray(img).astype(np.float32)
|
| | return float(calculate_niqe(arr, crop_border=0))
|
| |
|
| | def write_result_jsonl(file_path: str, data: dict):
|
| | """Append single result to file in JSONL format."""
|
| | try:
|
| | os.makedirs(os.path.dirname(file_path), exist_ok=True)
|
| | with open(file_path, 'a', encoding='utf-8') as f:
|
| | f.write(json.dumps(data, ensure_ascii=False, default=str) + '\n')
|
| | except Exception as e:
|
| | print(f"❌ Error writing JSONL file: {e}", file=sys.stderr)
|
| |
|
| | def main():
|
| | p = argparse.ArgumentParser(
|
| | description="Evaluate colorization/enhancement effect using CIEDE2000 and NIQE metrics, with JSONL output option")
|
| | p.add_argument("--groundtruth", type=str, required=True, help="Reference image URL or local path")
|
| | p.add_argument("--output", type=str, required=True, help="Reconstructed image URL or local path")
|
| | p.add_argument("--ciede-thresh", type=float, required=True,
|
| | help="CIEDE2000 minimum acceptance threshold (higher is better)")
|
| | p.add_argument("--niqe-thresh", type=float, required=True,
|
| | help="NIQE maximum acceptance threshold (lower is better)")
|
| | p.add_argument("--result", help="File path to store JSONL results")
|
| | args = p.parse_args()
|
| |
|
| | process_ok = True
|
| | comments = []
|
| |
|
| |
|
| | time_point = datetime.now().isoformat()
|
| |
|
| |
|
| | try:
|
| | img_ref = download_image(args.groundtruth)
|
| | img_recon = download_image(args.output)
|
| | except ValueError as err:
|
| | comments.append(str(err))
|
| | process_ok = False
|
| |
|
| |
|
| | if process_ok:
|
| | try:
|
| | score_ciede = compute_ciede2000(img_ref, img_recon)
|
| | score_niqe = compute_niqe(img_recon)
|
| |
|
| | comments.append(f"CIEDE2000 average color difference: {score_ciede:.4f} (threshold {args.ciede_thresh})")
|
| | comments.append(f"Reconstructed image NIQE score: {score_niqe:.4f} (threshold {args.niqe_thresh})")
|
| |
|
| | ok_ciede = score_ciede >= args.ciede_thresh
|
| | ok_niqe = score_niqe <= args.niqe_thresh
|
| |
|
| | if ok_ciede and ok_niqe:
|
| | comments.append("✅ Processing effect meets requirements: CIEDE2000↑ and NIQE↓ both satisfy thresholds")
|
| | result_ok = True
|
| | else:
|
| | fail_reasons = []
|
| | if not ok_ciede: fail_reasons.append("CIEDE2000 not met")
|
| | if not ok_niqe: fail_reasons.append("NIQE not met")
|
| | comments.append("❌ Processing effect does not meet requirements: " + " ".join(fail_reasons))
|
| | result_ok = False
|
| | except Exception as e:
|
| | comments.append(f"Exception during metric calculation: {e}")
|
| | result_ok = False
|
| | else:
|
| | result_ok = False
|
| |
|
| |
|
| | if args.result:
|
| | record = {
|
| | "Process": process_ok,
|
| | "Result": result_ok,
|
| | "TimePoint": time_point,
|
| | "comments": "\n".join(comments)
|
| | }
|
| | write_result_jsonl(args.result, record)
|
| |
|
| |
|
| | for line in comments:
|
| | print(line, file=(sys.stderr if not process_ok or not result_ok else sys.stdout))
|
| |
|
| | if __name__ == "__main__":
|
| | main() |
