Scribd
Upload
123 views9 pages

Understanding Structural Similarity Index (SSIM)

The SSIM index is an image quality metric that assesses luminance, contrast, and structure. It is calculated using local means, standard deviations, and cross-covariance of the original and distorted image. In Matlab, ssimval = ssim(A, ref) returns the SSIM index value between 0-1, where a value closer to 1 indicates better quality. SSIM provides a better measure of image quality compared to mean squared error and is more consistent with perceived visual quality.

Uploaded by

bob
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
123 views9 pages

Understanding Structural Similarity Index (SSIM)

The SSIM index is an image quality metric that assesses luminance, contrast, and structure. It is calculated using local means, standard deviations, and cross-covariance of the original and distorted image. In Matlab, ssimval = ssim(A, ref) returns the SSIM index value between 0-1, where a value closer to 1 indicates better quality. SSIM provides a better measure of image quality compared to mean squared error and is more consistent with perceived visual quality.

Uploaded by

bob
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 9

SSIM

Structural Similarity Index Measure

1
Explanation
• An image quality metric that assesses the visual impact of three characteristics of an
image: luminance, contrast and structure.

• The SSIM Index quality assessment index is based on the computation of three
terms, namely the luminance term, the contrast term and the structural term. The
overall index is a multiplicative combination of the three terms.

2
• μx, μy, σx,σy, and σxy are the local
means, standard deviations, and
cross-covariance for images x, y.

3
In Matlab: ssimval = ssim(A,ref)
• ssimval = ssim(A,ref) calculates the structural similarity (SSIM) index for grayscale image or
volume A using ref as the reference image or volume. A value closer to 1 indicates better image quality.
• A : Image for quality measurement:
Image for quality measurement, specified as a numeric array or a dlarray (Deep Learning Toolbox) object. If A is
not a 2-D grayscale image or 3-D grayscale volume, such as an RGB image or stack of grayscale images, specify
the DataFormat name-value argument. Do not specify the DataFormat name-value argument if A is a
formatted dlarray object.
• Ref : Reference image:
Reference image against which to measure quality, specified as a numeric array or a  dlarray (Deep Learning
Toolbox) object of the same size and data type as A. If ref is not a 2-D grayscale image or 3-D grayscale volume,
such as an RGB image or stack of grayscale images, specify the DataFormat name-value argument. Do not specify
the DataFormat name-value argument if ref is a formatted dlarray object.
• ssimval : SSIM index:
Numeric scalar with a single SSIM measurement.

4
Example 1

[ssimval, ssimmap] = ssim(A,ref)


5
Example 2
In the following, all distorted
images have roughly the same
mean squared error (MSE) values
with respect to the original
image, but very different quality.
SSIM gives a much better
indication of image quality.

6
Example 3
• The following example demonstrates the SSIM index map, which provides a
measurement of local image quality over space.

7
Example 4
• since the structural information of the image details are nearly completely lost, they
exhibit poorer visual quality. Close piece-by-piece comparison of the SSIM index and
the absolute error maps, we observe that the SSIM index is more consistent with
perceived quality measurement

8
• The scatter plots of the subjective measurement (mean opinion score, MOS) versus the
objective predictions (PSNR and MSSIM) are shown below, where each point represent
one test image.
• Clearly, MSSIM is much better in predicting the perceived image quality.

• PSNR: Peak Signal to Noise Ratio


• MSSIM: Mean Structural Similarity Index Measure
9

You might also like