Andrey Rubshtein
• Member for 10 years, 1 month
• Last seen more than a week ago

I would recommend a grid transform. Each point of the grid has a $(dx,dy)$ offset, and a point in the middle of a square is interpolated according to the control points. Specifically in your case, you ...

Regarding LAB, it is a good way if you are interested in the differences as humans perceive them. About texture, I would suggest taking a look at some proprietary texture descriptors: Gray level co-...

You can use any target you'd like as long as you can find the features in a robust way, and you know their location in the real world. For example I used once ISO12233 chart image to calibrate radial ...

The images you are working with are probably with gamma compensation, you should remove it first by applying inverse gamma. More accurately, each camera has its own tone curve mapping. You can build a ...

Another good texture measure would be gray-level co-occurence matrix, which is implemented in Matlab as graycomatrix. By the way, in your case a simple Otsu threshold might do the trick as well.

You should use the function maketform to create your homography, and apply it on image by using imtransform. Don't forget to apply transpose on your matrix, because Matlab works on pixels as rows, not ...

You could check how "circular" a blob is, by computing the ratio between area and the square of perimeter. For a circle that would be: $A = \pi r^2 , P = 2 \pi r$, which implies that $\frac{A}{P^... View answer Accepted answer 2 votes You should take a photograph of an object with known color values, such as Macbeth chart: By using the lowest row of grayscale colors, you will be able to find out the tone-mapping (sometimes ... View answer 2 votes Assume that the center of the square maps to the center of the vertexes. The vertices in the output are known. Assume they are$(x_1,y_1),...,(x_4,y_4) $Their center is$ ( \frac{x_1+x_2+x_3+x_4}{4},\...

Yes it is. That what's Kalman filter is all about. Your measurement includes position data, and there is a hidden velocity parameter. You will need some kind of model, like constant-velocity. You can ...

If the lines are not parallel, you can calculate the point of their intersection and use it as a point of reference. In your painting, you can use the purple points as well: By the way, the ...

Here is a possible solution: Use FFT2 to find angle of text and approximate spaces between lines Rotate the image to be at 0 angle Sum the image column-wise Find threshold that splits lines of text ...

The logo you are looking at is symmetric. For example, you could rotate it by 120 degrees and get the same pattern. If you take a look, many of the "errors" are not really errors. The same pattern ...

Assuming that the image is gray level (not color) Use Otsu's thresholding method to detect whether the image has bi-modal histogram. If Otsu tells you with high confidence that the image is bi-modal, ...

The first part of the problem that you're describing is called registration. Assuming that the displacement is indeed small, and your problem has no time limits, you can use normalized cross-...

You could solve the problem in several ways: Image pre-filtering (smoothing) with some kind of Gaussian filter, before doing edge detection, will yield smoother edges. Morphological operations, such ...

Let's think of the ideal descriptor to understand the idea. An ideal descriptor is a function from pixel space to some other space, such that same objects have the same outcome, whereas different ...

Assuming that there is a ground truth, (at least theoretically) one of the possible ways to overcome the "tediousness" problem is a "bootstrap" ground truth creation. If you already have a decent ...

Basically, it is impossible to calculate the full DFT of your data. Take a look at the DFT matrix. All of its elements are non-zeroes. Thus, each and every signal element is important for any ...

Another good explanation is that you enlarge the amount of energy in your image, and then subtract only the low frequencies from your image, thus making the high frequencies more dominant: 4 * [0 1 ...

Most of the image processing/computer vision algorithms are ad hoc. Your algorithm is good. However, I think that a few points can be improved: Instead of finding median color, you could do use the ...

You should photograph a known calibration target with your system, (like this one or the one that I put in the thumbnail). This will allow you discovering the OECF of your system. You can either ...

What you need is an interpolation method. The method you described is called nearest-neighbor, because you pick the pixel that is nearest to the place you actually wanted. Other methods include: ...

You should do it on a known set of calibration targets. For example, this type of calibration target is being used to measure noise. You should find the patches locations, and measure the noise by ...

In case your filter is linear, this should be quite easy. You have a linear model with unknown parameters (9 if your support is 3x3, 16 if 4x4, etc...). You have a lot of constraints - the original ...

Though the question is classic signal processing, you can use a computer vision technique called Hough line detector. The idea is that each pixel votes for all the possible lines that pass through it. ...

The convolution operation is associative. $(S ** g) ** h = S ** ( g ** h )$ This means that if you find two kernels $g,h$ such that $g**h = k$, (where $k$ is your kernel), you can split your ...

That depends on the definition of high-pass filter. If you define a high-pass filter as a filter that has high response in the high frequencies in frequency domain, then the easiest way is to take a ...