How do I lower an image quality?

I'm new to matlab and need to create a for loop that gradually lowers the quality of an image in a GUI. I tried searching on the internet but can't find how to do it. I'm aware it's probably imwrite but I can't get it to work properly.

The image is .jpeg format, I just need to know how to lower it once and I'll figure everything else from there.

[Mann Baidi]

What do you mean by 'lower an image quality'?

Does it means changing the resloution of the image or degrade the quality by adding some noise in it?

Answer 1

[Harsh]

Hi,

It seems like you are interested in learning how to modify the resolution of an image and compress an image using MATLAB. Here are simplified steps for doing both:

1. To change the resolution of an image:

Use the imresize function. This function allows you to adjust the image's size to a new resolution.

image = imread("SampleImage.jpeg"); 
lowResImage = imresize(image, 0.5); 

2. To compress an image:

Use the wcompress function. This function is designed for image compression.

As an example on image compression using wcompress, you can open the "Image Compression Using Basic Parameters" example in MATLAB with the following command:

openExample('wavelet/ImageCompressionUsingBasicParametersExample') 

Refer to the following links to gather more information about the "imresize" and "wcompress".

1. https://www.mathworks.com/help/matlab/ref/imresize.html
2. https://www.mathworks.com/help/wavelet/ref/wcompress.html

I hope this helps, thanks!

[Sebastián López Rendón]

Thank you!

Answer 2

[John D'Errico]

There are other ways to lower image quality. Several ideas have already been mentioned.

1. Resize the image, then expand it back up again to the original size. Effectively, that adds in artifacts due to the interpolation needed to resize the image.
2. Compress the image, then decompress it. And image compression can be done in a lossless way, but if more compression is done, then you will incur image artifacts due to the compression.
3. Apply noise to the image. The amount of noise will influence the degree of quality loss.
4. Apply a blur of some sort. This might be a motion blur. It might be a general gaussian smear.

Each of these image degradations will have characteristics of their own. I'm sure we could think of a few more ways to achieve a quality degradation, as image quality has many facets. You will first need to decide what source of quality loss seems appropriate to you.

[Image Analyst]

Just look at the tons of filters that you can apply in Adobe Photoshop. For example, turn the image into a painting or stained glass window or pixelize it. Each of them changes the image in some way and that could be interpreted as lowering the quality of the image as compared to the original.

[DGM]

It could be as simple as a change of contrast obscuring key features. Without defined goals, it's as bad as words like "enhance".

Answer 3

[Sarthak]

Hi Sebastian,

As you suggested, you can indeed use imwrite to reduce the quality of an image using the 'Quality' argument. For more information on how to use it, please refer to the link below:

https://www.mathworks.com/help/matlab/ref/imwrite.html#mw_0762e49f-e3b0-4b66-bcfb-0ef235680b4f

You can create your 'for' loop iterating over quality with your desired step size and use the following line of code in your 'for' loop for gradually lowering the quality of the image.

imwrite(originalImage, outputFileName, 'Quality', quality);

Hope this helps!

[DGM]

• That only applies to JPEG outputs.
• That's a very specific definition of "quality".
• The resultant (and change in) "quality" depends significantly on the source image.
• This can only reduce the quality of the image, even for a 100% quality setting.

All standard JPEG files written by MATLAB are 4:2:0 chroma subsampled. They're all degraded images.