Tetragramm's profile - activity

You are correct. +Z should be in front of the camera. Not sure what I was thinking.

Try shining an incandescent light on the paper. That should help distinguish the light and dark parts, though I don't k

Unhandled exception at 0x00007FFB8B3F9159 in test1.exe: Microsoft C++ exception: cv::Exception at memory location 0x0000

EDIT: See comments about saturate_cast and the missing else. Also, that is what references are for. Passing things int

Oh, I see. saturate_cast will do what needs to be done, but your problem is your if(iVal<0). If it is less than 0,

cv::multiply is not a cross product. It is an element-wise product. You want to do vec_cross_result = PQ.cross(PR);

Did you try that exact line?

int iVal = inputImage1->at<uchar>(i,j) - inputImage2->at<uchar>(i,j); This line. Cast them to int be

int iVal = inputImage1->at<uchar>(i,j) - inputImage2->at<uchar>(i,j); This line. Cast them to int be

You can extract the modeled background from the OpenCV BackgroundSubtractors. https://docs.opencv.org/4.2.0/d7/df6/clas

Per the documentation, see the papers [271] Zoran Zivkovic and Ferdinand van der Heijden. Efficient adaptive density es

I'm not sure, but does the array have 4 channels? If so, try setting the last one (which may be treated as a transparen

Well, if these are representative images, you are probably better off doing something besides HoughCircles. I suggest d

That sounds like the CV_32F contains a NaN to begin with. Try running patchNaNs before mean to see if that fixes it.

I would like to think I had a reason for putting that there, but I can't remember, and it certainly looks unnecessary.

Reading is no problem at all. Writing to different locations is ok, as long as you don't do anything that alters the Ma

https://www.carrida-technologies.com/doc/SDK/4.3.1/html/classsvcapture_1_1_image.html Well that's remarkably un-informa

Are the only values 0, 127 and 255?

It is much more accurate to get the depth at the particular pixel. That accounts for all the error in whether the table

They have depth. You do not. If you know the distance to the table (t), and the table is parallel to the plane(R), y

The short answer is, you can't do this properly, but you can fake it. Long answer is, because your camera isn't aligned

Pretty hard. However, you can build and link the static libraries instead of DLLs, and it should, maybe, possibly remov

Perhaps you could share the code you used to call undistortPoints? I can't possibly tell if you're using it correctly.

When the camera is parallel to the table, yes, you can make the assumption that R is zero. t must obviously be [0,0,z],

Then do please make a pull-request on github.

Both. This uses some known guarantees about the Mat memory structure to be much faster than a pixel-by-pixel method.

Nope, that's not the right calculation. And yes, it could also be the non-flatness of the table. But unless that slope

There are two main possibilities. It could be either, or both. First is camera distortion. Distortion typically gets

If your object_points are described in mm, then your tvec will similarly be in mm.

First a question. You wish to move to an X,Y. Is it true that the farther from the destination you are, the farther th

Well, you can make the change for your personal use and submit it as a pull-request. But PowerPC is not a particularly

The size of the binaries is a concern, and this is doubling the size of the code for basically all the cvtColors. There

That would work as long as the camera is pointed straight down at the table. And as long as you're ok with the center of

This sort of question is best answered by the documentation for the function HERE. calibrateCamera treats the corner of

Blob Detector Not working when it should on obvious blobs, makes no sense So I have an HSV filtered image that I am tryi

You said you know the orientation and position of the camera. Is that from an opencv function that gives you rvec and t

Based on your clarification, you can do what you need. First you calculate the vector in space. If you don't have the

Take a look at cv::fisheye::undistortPoints.

Do you not need z coordinate because you know how far it is? Or at least the location of the camera relative to some su

OpenCV provides the ComposeRT function, which combines two sets of Rotation and Translation transforms.

Try a simple chessboard pattern. A 2x2 chessboard can be seen from very far away, and the center is the center no matter

Probably because you're asking for 3 uchar values. Look at the lines with .at<Vec3b>. The type between the <

So, I can see a lot more than one feature. Every corner is a feature. You have to do some logic to match those to the 3

You want to use cv::moments using a.col(i), where i is the column number. The output is a structure that contains the m

So my thought is this. You are using something with real cameras. They have lenses that look something like THIS. But

Are the cameras pointing straight, or are they canted inwards or outwards? IE: ||, /\, or \/ One possibility is that t