Skip to main content
Share Your Experience: Take the 2024 Developer Survey
19 votes
Accepted

Why should an image be blurred using a Gaussian Kernel before downsampling?

An image "should not be blurred using a Gaussian Kernel" in general. This however can be a safe bet for a lot of basic image processing needs, and a smoothing is almost mandatory when you ...
Laurent Duval's user avatar
19 votes
Accepted

What is different between Euclidean distance and the Geodesic distance?

Simple: the Euclidean distance completely ignores the shape when finding a path from the start point to the end point while, for the geodesic distance, the path is constrained to be within the given ...
Peter K.'s user avatar
  • 25.8k
18 votes
Accepted

What Is the Difference between Difference of Gaussian, Laplace of Gaussian, and Mexican Hat Wavelet?

Laplace of Gaussian The Laplace of Gaussian (LoG) of image $f$ can be written as $$ \nabla^2 (f * g) = f * \nabla^2 g $$ with $g$ the Gaussian kernel and $*$ the convolution. That is, the Laplace ...
Cris Luengo's user avatar
  • 2,584
18 votes

Why color is coded with two numbers rather than one?

Note: also check out the comments for more pointers. I'll assume that you think of wavelengths, or hue and saturation, which is why you think you could maybe represent color as a single value. You ...
Christoph Rackwitz's user avatar
17 votes
Accepted

Why real part of FFT converts image into rotation + original?

Let's say your image is given by $I(x,y)$. Then its Fourier transform is given by $$ I^f(\omega_x,\omega_y) = \int_x\int_yI(x,y)e^{j\omega_xx}e^{j\omega_yy}dxdy $$ Now you take the real part and ...
ThP's user avatar
  • 1,460
17 votes
Accepted

what does "derivative" means in image processing?

Look at the numbers in the filter kernel in just 1 dimension for simplicity. For a Sobel and Prewitt matrix you have something that roughly looks like this ...
Atul Ingle's user avatar
  • 4,134
16 votes
Accepted

Which image filter can be applied to remove gridded pattern from corrupt jpegs?

You can use a standard inpainting algorithm. These algorithms replace marked pixels in an image with the pixel values that surround these marked pixels. The challenge here is to detect the grid (my ...
Maximilian Matthé's user avatar
16 votes
Accepted

DSP or signal/image/data processing jokes

An airplane is leaving Warsaw (the capital of Poland), and gets caught in a terrible winter storm. The plane rolls, pitches and yaws. The crew is expecting the plane to crash or break up at any time. ...
Laurent Duval's user avatar
15 votes
Accepted

Does a simple photograph contain more information than a complex painting?

It depends how you define the term "information" or "entropy". The conventional definition of entropy of an image is to think the image as a two-dimensional matrix of pixels and $$H = - \sum_k p_k \...
AlexTP's user avatar
  • 6,595
14 votes
Accepted

Why does this Moiré pattern look like this?

You'll need to understand the sampling theorem. In short, each signal has what we call a spectrum¹, which is the Fourier transform of the signal as it comes in time domain (if it is a time signal), or ...
Marcus Müller's user avatar
14 votes
Accepted

Auto Detection of Rotation Angle on Arbitrary Image with Orthogonal Features

If I understand your method 1 correctly, with it, if you used a circularly symmetrical region and did the rotation about the center of the region, you would eliminate the region's dependency on the ...
Olli Niemitalo's user avatar
13 votes

Is deep learning killing image processing/computer vision?

Today we had a discussion with a friend of mine. It was a rainy day here in Munich, while a large portion of Europe was having a kind of sunny atmosphere. People were sharing photographs in social ...
Tolga Birdal's user avatar
  • 5,465
11 votes

In opecv function Hough circles how does parameter 1 and 2 affect circle detection and how can I adjust them to increase accuracy?

Intuition for parameters of HoughCircles: image: 8-bit, single channel image. If working with a color image, convert to grayscale first. method: Defines the method to detect circles in images. ...
Eric Leschinski's user avatar
11 votes
Accepted

2D Convolution as a Doubly Block Circulant Matrix Operating on a Vector

The point is that circular convolution of two 1-D discrete signals can be expressed as the product of a circulant matrix and the vector representation of the other signal. The circulant matrix is a ...
msm's user avatar
  • 4,295
11 votes

Peak Signal to Noise Ratio (PSNR) in Python for an Image

You can use cv2.PSNR like this example: import cv2 img1 = cv2.imread('img1.bmp') img2 = cv2.imread('img2.bmp') psnr = cv2.PSNR(img1, img2)
Amir Khakpour's user avatar
11 votes
Accepted

Python vs Matlab? Which one better for image processing?

[Comparison with cars at the end] For learning from scratch, I would not suggest a programming language alone, but instead the couple "teaching materials" (book, lecture notes) + "...
Laurent Duval's user avatar
10 votes

How do I get the most accurate camera calibration?

I decided to post this answer here because a while back, this came up as the top result in Google and its suggestions helped me. So I decided to share my experience too. Having spent countless hours ...
Maghoumi's user avatar
  • 201
10 votes
Accepted

When is a network called end-to-end training?

From feature extraction to learning the desired result, deep learning algorithms can act as full pipelines for solving tasks at hand. End-to-end learning usually refers to omitting any hand-crafted ...
Tolga Birdal's user avatar
  • 5,465
10 votes
Accepted

Subpixel - what is it?

Images are technically limited to the original pixel resolution. Objects in an image are spatially quantized at this resolution. However, an edge for instance may be localized "between" native pixels. ...
Laurent Duval's user avatar
10 votes
Accepted

Why does the separable filter reduce the cost of computing the operator?

Assume you have a $N\times M$ sized image. If you know take what is classically used, a square filter kernel, of let's say size $L\times L$, you'd need to convolve that with the picture – which gives ...
Marcus Müller's user avatar
10 votes

DSP or signal/image/data processing jokes

In high school math class I finally got the courage up to ask the cute and brainy girl at the front row out on a date. She just looked at me with a discouraging face and as she waved her hand face ...
Dan Boschen's user avatar
10 votes
Accepted

Which Noise Reduction Algorithms Are Used in Commercial RAW Image Processors?

Common Approaches for Commercial Denoisers Commercial denoisers are different than what you'd see on most papers. While on papers the results are mostly using objective metrics (PSNR / SSIM) and are ...
Royi's user avatar
  • 19.7k
10 votes
Accepted

Pros, Cons, and Applications of SVD-based Image Compression (Singular Value Decomposition)

The idea of using PCA with images is to look at the margin ensembles generated by the image. That means, we consider two distributions: The distribution of the column over the realisations of the rows....
Jazzmaniac's user avatar
  • 4,583
9 votes

Best way of segmenting veins in leaves?

Following on from the above excellent answer, here is how to do it in python using scikit funcitons. ...
Matthew Shun-Shin's user avatar
9 votes
Accepted

How to calculate particles sizes for spheric particles with overlapping and superposition? (Example image included)

Here is what I experimented with: Use ELSD to generate elliptic contours. You could basically use any edge detector, but since in the following stages I will benefit from circle detectors, it is good ...
Tolga Birdal's user avatar
  • 5,465
9 votes

Why real part of FFT converts image into rotation + original?

The result ThP provided can also be stated in very simple terms: If you have a data set which is purely real, its (inverse) Fourier transform will have Hermitian symmetry: If you find the value $z$ at ...
M529's user avatar
  • 1,736
9 votes

DSP or signal/image/data processing jokes

Crazy Audio Fx ilustration lol more:
ederwander's user avatar
  • 2,038
9 votes

Why should an image be blurred using a Gaussian Kernel before downsampling?

According to (digital) sampling theorem, signals should be properly bandlimited, before they are (down) sampled. A practical digital filter approximately limits the bandwidth of the signal and makes ...
Fat32's user avatar
  • 28.3k
9 votes
Accepted

Image processing vs Computer vision?

Digital image processing is an extension of digital signal processing and linear system theory into two dimensional signals. Image processing involves all low level tasks such as filter design and ...
Fat32's user avatar
  • 28.3k

Only top scored, non community-wiki answers of a minimum length are eligible