AbdurRasheed Sadiq asked . 2023-02-11

Why do I find different SSIM values depending on whether the images

Why do I find different SSIM values depending on whether the images' intensity information is stored in the range [0 1] or [0 255]?

I import an image using the im2gray() function, which stores it as an array with a range of values between 0 and 255. I convert it into a double using double(), which does not alter the values of the elements but just changes the file format. I then alter the values and compare the altered image to the original in two ways. Firstly, I use mat2gray() to convert the intensity information in both images to the range 0-1 instead of 0-255. Secondly, I compare the 0-255 matrices. The SSIM values I get are quite different (the first value is higher than the second).

Why is this? Does the mat2gray() function lose image information? Which is the "correct" value of SSIM between the original and distorted images?

NB you can use any image for this. The one I used it attached to this answer.

% Import an image using im2gray
im = im2gray(imread('peppers.png'));

% Convert this a double (without rescaling the values from the range 0-255)
% and then modify the values slightly to create a distorted image.

imdouble = double(im);
imdouble_new = imdouble * 0.5;
im_new = mat2gray(imdouble_new, [0 255]);

% NB im_new is a grayscale image with values 0-1 rather than 0-255. If we
% want to compare this image to the original using the SSIM metric we can
% do so in 2 ways.

% 1) Convert im into a 0-1 image matrix and use the ssim function.

ssim_values_1 = ssim(mat2gray(imdouble, [0 255]), im_new);
ssim_values_1(1)

% 2) Just compare the double matrices using the ssim function.

ssim_values_2 = ssim(imdouble, imdouble_new);
ssim_values_2(1)

Ans = 0.7508

% 2) Just compare the double matrices using the ssim function.
ssim_values_2 = ssim(imdouble, imdouble_new);
ssim_values_2(1)

ans = 0.6400

intensity-information

image processing , image analysis , ssim , grayscale , Image Segmentation and Analysis , Image Qual

Expert Answer

Neeta Dsouza answered . 2024-12-21 17:23:43

You're passing images to ssim() which are improperly-scaled for their class. Many functions make assumptions about the expected levels of an image based on its class, and ssim() is one. It uses the class to implicitly select the dynamic range parameter. The unit-scale ([0 1]) images work fine, since that's what ssim() expects a floating point image to be. It's the uint8-scale ([0 255]) images that it doesn't expect.

You can either cast the uint8-scale images as uint8, or to avoid the rounding loss on a working image, you can explicitly set the dynamic range parameter:

% Import an image using im2gray
im = rgb2gray(imread('peppers.png'));

% Convert this a double (without rescaling the values from the range 0-255)
% and then modify the values slightly to create a distorted image.
imuint8 = double(im);
imuint8_half = imuint8 * 0.5;

imdouble = mat2gray(imuint8, [0 255]);
imdouble_half = mat2gray(imuint8_half, [0 255]);

% NB im_new is a grayscale image with values 0-1 rather than 0-255. If we
% want to compare this image to the original using the SSIM metric we can
% do so in 2 ways.

% 1) Convert im into a 0-1 image matrix and use the ssim function.
ssim_values_1 = ssim(imdouble, imdouble_half,'dynamicrange',1);
ssim_values_1(1)

% 2) Just compare the uint8-scale matrices using the ssim function.
ssim_values_2 = ssim(imuint8, imuint8_half,'dynamicrange',255);
ssim_values_2(1)

ans = 0.7508

% 2) Just compare the uint8-scale matrices using the ssim function.
ssim_values_2 = ssim(imuint8, imuint8_half,'dynamicrange',255);
ssim_values_2(1)

ans = 0.7508

Note that I renamed the variables so that their names describe to their data range.

Not satisfied with the answer ?? ASK NOW

Recent Matlab Projects

Classification of Covid and Non-Covid Lungs CT-Scan using Deep Learning with MATLAB

Matlab simulation on Wind Energy system

Frequently Asked Questions

MATLAB offers tools for real-time AI applications, including Simulink for modeling and simulation. It can be used for developing algorithms and control systems for autonomous vehicles, robots, and other real-time AI systems.

MATLAB Online™ provides access to MATLAB® from your web browser. With MATLAB Online, your files are stored on MATLAB Drive™ and are available wherever you go. MATLAB Drive Connector synchronizes your files between your computers and MATLAB Online, providing offline access and eliminating the need to manually upload or download files. You can also run your files from the convenience of your smartphone or tablet by connecting to MathWorks® Cloud through the MATLAB Mobile™ app.

Yes, MATLAB provides tools and frameworks for deep learning, including the Deep Learning Toolbox. You can use MATLAB for tasks like building and training neural networks, image classification, and natural language processing.

MATLAB and Python are both popular choices for AI development. MATLAB is known for its ease of use in mathematical computations and its extensive toolbox for AI and machine learning. Python, on the other hand, has a vast ecosystem of libraries like TensorFlow and PyTorch. The choice depends on your preferences and project requirements.

You can find support, discussion forums, and a community of MATLAB users on the MATLAB website, Matlansolutions forums, and other AI-related online communities. Remember that MATLAB's capabilities in AI and machine learning continue to evolve, so staying updated with the latest features and resources is essential for effective AI development using MATLAB.

Without any hesitation the answer to this question is NO. The service we offer is 100% legal, legitimate and won't make you a cheater. Read and discover exactly what an essay writing service is and how when used correctly, is a valuable teaching aid and no more akin to cheating than a tutor's 'model essay' or the many published essay guides available from your local book shop. You should use the work as a reference and should not hand over the exact copy of it.