Introduction:

In the realm of computer vision, where algorithms strive to make sense of the visual world, the presence of noise can be akin to a discordant note in an otherwise harmonious melody. Noise, in the context of computer vision, refers to unwanted or random variations in pixel values that can distort or degrade the quality of an image. Understanding the various types of noise is crucial for developing robust computer vision systems that can accurately interpret visual information. In this blog, we'll explore some common types of noise that often plague digital images and how they can be mitigated.

So if images are just functions, then we can do things to images that we can do to functions. Like we can just add them, right? You can add two functions, right? Well, then we can add two images.

Section 1- What is noise in the image? 

And to introduce this a little bit, we're going to introduce the concept of noise, okay? So noise in an image. Is just another function that, combined with the original image, gives us a new function. So, we'll just write this, this way as our new image. We'll call it I Prime. It's just I of x, y plus this noise function. You know, well what does that mean? Well, we have to take a look at what this noise function would be. 

Common Type of Noise Function 

Okay, so there are lots of different kinds of noise functions.

1- Salt and pepper Noise 

Salt and pepper noise, also known as impulse noise, is a type of noise that appears as random white and black pixels in an image. It is caused by sharp and sudden disturbances in the image signa

Contrary to the subtlety of Gaussian noise, salt-and-pepper noise manifests as randomly occurring white and black pixels scattered throughout an image. This type of noise is common in images acquired in low-light conditions or transmitted over noisy channels. Median filtering, a non-linear filtering technique, is commonly employed to suppress salt-and-pepper noise, effectively removing isolated outliers without blurring the overall image.

Here's one, and this stuff's courtesy of Steve Sites, there's a type of noise called salt and pepper noise. Which doesn't take a rocket scientist for you to figure out that probably what it does is, it takes your original picture and it sprinkles occasional white spots and occasional dark spots. And that's called salt and pepper noise for the, for the obvious reason.

 2- Impulse noise

A, relative to that is something called an impulse noise, where you just get little white specks now and then. Different kind of imaging systems might give you that kind of noise.

3- Gaussian  noise

Imagine a gentle sprinkle of random fluctuations across an image – that's Gaussian noise. It is characterized by pixel values varying according to a Gaussian distribution. This type of noise is prevalent in real-world images due to factors like sensor imperfections, transmission errors, or electronic interference. Gaussian noise is often addressed using smoothing filters like Gaussian filters or bilateral filters, which help in reducing the impact of these random fluctuations.

But by far, the noise that you're most familiar with is typically Gaussian noise, or normally distributed noise. Where we basically assume that at every pixel we take the original image and we stick on here some value that is independent identically distributed from some normal or some Gaussian distribution. All right, and that's Gaussian noise. And most of the time when we talk about noise we'll talk about that function. 

Okay? We can actually have Matlab make us a noise function. It's real easy. So here we say, look we're going to make a noise array, which is just, I take the size of my image, random n, randn generates a noise signal that has a mean of zero and a standard deviation of one, and if we scale that up by some sigma. Okay? That will spread that out and make it bigger so that's essentially the noise with mean of zero and a, a standard deviation of sigma. And because functions are just functions and images are functions, I can just add them. I can say let my output just be the image plus the noise. And if I were to plot that, you would see what's here, right? And on the right you can see that there's all this noise in our peppers. And if we plot this, you can see here we get this nice clean plot and here we have all this extra noise that's been added. And so that's our noise function.

4- Quantization Noise:

When an image is digitized, continuous-tone values are converted into discrete digital values, introducing quantization noise. This noise is a consequence of the limited precision of pixel values, leading to rounding errors. Increasing the bit-depth during image acquisition can help reduce quantization noise. In applications where precision is critical, higher bit-depth sensors or floating-point representations can be employed.

5- Shot Noise:

Arising from the statistical nature of light, shot noise is inherent in any imaging system. It is particularly noticeable in low-light conditions where the number of photons striking the image sensor is low. Shot noise can be reduced by increasing the exposure time, using larger apertures, or employing image post-processing techniques.

6- Temporal Noise

In video sequences or real-time applications, temporal noise is a concern. It manifests as flickering or variations in pixel values over consecutive frames. This can be caused by fluctuations in lighting conditions, sensor instabilities, or electronic noise. Temporal filtering techniques, such as temporal averaging or using specialized sensors with improved temporal stability, can help mitigate this type of noise.

7- Color Noise

In color images, each channel (Red, Green, and Blue) can be affected by noise independently. Color noise can manifest as unwanted variations in color intensity. Techniques like channel-wise filtering or color space transformations can be employed to address color noise while preserving the overall color balance.

Conclusion:

As computer vision continues to evolve and find applications in diverse fields, the ability to handle and mitigate various types of noise becomes paramount. Researchers and engineers in the field continually innovate new algorithms and techniques to enhance the robustness and accuracy of computer vision systems. By understanding the nuances of different noise types and employing appropriate noise reduction strategies, we pave the way for clearer, more reliable, and more accurate computer vision applications in the real world.