niqe

Naturalness Image Quality Evaluator (NIQE) no-reference image quality score

Syntax

score = niqe(A)

score = niqe(A,model)

Description

score = niqe(A) calculates the no-reference image quality score for image A using the Naturalness Image Quality Evaluator (NIQE). niqe compares A to a default model computed from images of natural scenes. A smaller score indicates better perceptual quality.

score = niqe(A,model) calculates the image quality score using a custom model.

Examples

Calculate NIQE Score Using Default Feature Model

Compute the NIQE score for a natural image and its distorted versions using the default model.

Read an image into the workspace. Create copies of the image with noise and blurring distortions.

I = imread('lighthouse.png');
Inoise = imnoise(I,'salt & pepper',0.02);
Iblur = imgaussfilt(I,2);

Display the images.

figure
montage({I,Inoise,Iblur},'Size',[1 3])
title('Original Image | Noisy Image | Blurry Image')

Calculate the NIQE score for each image using the default model. Display the score.

niqeI = niqe(I);
fprintf('NIQE score for original image is %0.4f.\n',niqeI)

NIQE score for original image is 2.5455.

niqeInoise = niqe(Inoise);
fprintf('NIQE score for noisy image is %0.4f.\n',niqeInoise)

NIQE score for noisy image is 10.8770.

niqeIblur = niqe(Iblur);
fprintf('NIQE score for blurry image is %0.4f.\n',niqeIblur)

NIQE score for blurry image is 5.2661.

The original undistorted image has the best perceptual quality and therefore the lowest NIQE score.

Calculate NIQE Score Using Custom Feature Model

Load a set of natural images into an image datastore. These images are shipped in Image Processing Toolbox™ in a directory named 'imdata'.

setDir = fullfile(toolboxdir('images'),'imdata');
imds = imageDatastore(setDir,'FileExtensions',{'.jpg'});

Train a custom NIQE model using the image datastore.

model = fitniqe(imds);

Extracting features from 37 images.

Completed 2 of 37 images.  Time: Calculating...

Completed 3 of 37 images.  Time: 00:27 of 07:29

Completed 4 of 37 images.  Time: 00:39 of 07:26
..
Completed 11 of 37 images.  Time: 00:50 of 02:32

Completed 12 of 37 images.  Time: 01:01 of 02:57
.
Completed 14 of 37 images.  Time: 01:20 of 03:25
.
Completed 21 of 37 images.  Time: 01:31 of 02:35
..
Completed 24 of 37 images.  Time: 01:46 of 02:40
.....
Completed 35 of 37 images.  Time: 01:57 of 02:03
.
Done.

Read an image of a natural scene. Display the image.

I = imread('car1.jpg');
imshow(I)

Calculate the NIQE score for the image using the custom model. Display the score.

niqeI = niqe(I,model);
fprintf('NIQE score for the image is %0.4f.\n',niqeI)

NIQE score for the image is 1.8601.

Input Arguments

A — Input image

2-D grayscale image | 2-D RGB image

Input image, specified as a 2-D grayscale or RGB image.

Data Types: single | double | int16 | uint8 | uint16

model — Custom model

niqeModel object

Custom model of image features, specified as a niqeModel object. model is derived from natural scene statistics.

Output Arguments

score — No-reference image quality score

nonnegative scalar

No-reference image quality score, returned as a nonnegative scalar. Lower values of score reflect better perceptual quality of image A with respect to the input model.

Data Types: double

Algorithms

NIQE measures the distance between the NSS-based features calculated from image A to the features obtained from an image database used to train the model. The features are modeled as multidimensional Gaussian distributions.

References

[1] Mittal, A., R. Soundararajan, and A. C. Bovik. "Making a Completely Blind Image Quality Analyzer." IEEE Signal Processing Letters. Vol. 22, Number 3, March 2013, pp. 209–212.