baidut / matLIVE

Matlab code collection of LIVE-IQA

Mat

Matlab library for Image/Video Quality Assessment.

For python library, please check out PyLIVE

Please go to lab recourse page for more resources.

Installation

Add all folders to path, then you're all set 😃

Usage

Please cite the corresponding papers in any published work if you use these code. Bibtex entries are provided here.

• SSIM(img1, img2) structural similarity (SSIM) index \cite{wang2004SSIM} ☻
• MSSSIM(img1, img2) multi-scaleSSIM (MS-SSIM) index \cite{wang2003MSSSIM} ☻
• IFC(distorted, pristine) information fidelity criterion (IFC) index \cite{sheikh2005IFC} ☻✚ 📁 Evaluation of IFC
• VIF(distorted, pristine) visual information fidelity (VIF) index \cite{sheikh2004VIF} ☻✚ 📁
• VIFP(distorted, pristine) pixel-based VIF \cite{sheikh2004VIF} ☻
• VSNR(distorted, pristine) visual signal to noise ratio (VSNR) \cite{chandler2007VSNR} ☻ ✚:file_folder:
• MAD(distorted, pristine) The Most Apparent Distortion (MAD) \cite{larson2010MAD}

---

Unchecked items are under development.

The meaning of the icons is as follows:

☻ rgb2gray is applied for color images

✚ remember addpath

📁 require matlabPyrTools, remember add matlabPyrTools to your path. Matlab Pyramid Tools

Notes

• Some metrics are symmetric, e.g. SSIM(img1, img2) == SSIM(img2, img1) while some are not e.g. VIF(img1, img2) ~= VIF(img2, img1)
• When using IQAs to evaluate image enhancement algorithms, we suggest use the original image as distorted, since it has a worse quality than the enhanced one, e.g. VIF(original, enhanced)
• We use IQA(imdist,imorg) while some original implementation use IQA(imorg,imdist)

References

@article{wang2004SSIM,
  title={Image quality assessment: from error visibility to structural similarity},
  author={Wang, Zhou and Bovik, Alan C and Sheikh, Hamid R and Simoncelli, Eero P and others},
  journal={IEEE transactions on image processing},
  volume={13},
  number={4},
  pages={600--612},
  year={2004}
}
@inproceedings{wang2003MSSSIM,
  title={Multiscale structural similarity for image quality assessment},
  author={Wang, Zhou and Simoncelli, Eero P and Bovik, Alan C},
  booktitle={The Thrity-Seventh Asilomar Conference on Signals, Systems \& Computers, 2003},
  volume={2},
  pages={1398--1402},
  year={2003},
  organization={Ieee}
}
@article{sheikh2005IFC,
  title={An information fidelity criterion for image quality assessment using natural scene statistics},
  author={Sheikh, Hamid R and Bovik, Alan C and De Veciana, Gustavo},
  journal={IEEE Transactions on image processing},
  volume={14},
  number={12},
  pages={2117--2128},
  year={2005},
  publisher={IEEE}
}
@inproceedings{sheikh2004VIF,
  title={Image information and visual quality},
  author={Sheikh, Hamid R and Bovik, Alan C},
  booktitle={2004 IEEE International Conference on Acoustics, Speech, and Signal Processing},
  volume={3},
  pages={iii--709},
  year={2004},
  organization={IEEE}
}
@article{chandler2007VSNR,
  title={VSNR: A wavelet-based visual signal-to-noise ratio for natural images},
  author={Chandler, Damon M and Hemami, Sheila S},
  journal={IEEE transactions on image processing},
  volume={16},
  number={9},
  pages={2284--2298},
  year={2007},
  publisher={IEEE}
}
@article{larson2010MAD,
  title={Most apparent distortion: full-reference image quality assessment and the role of strategy},
  author={Larson, Eric Cooper and Chandler, Damon Michael},
  journal={Journal of Electronic Imaging},
  volume={19},
  number={1},
  pages={011006},
  year={2010},
  publisher={International Society for Optics and Photonics}
}

Log

• VIF - case sensitive, change vifsub_est_m to vifsub_est_M