swebb_denver's profile - activity

I hope that all makes sense. I'm pretty sure it would do what you want, but maybe someone else has a better approach.

It's hard to follow the code above because the carriage returns are ignored for whatever reason. the main point of the

It's hard to follow the code above because the carriage returns are ignored for whatever reason. the main point of the

So my arCamMat was estimated, but close to actual. I ignore the rvec/tvec return values, and I was only interested in

If that all makes sense, the key OpenCV call that I made was this: // The purpose of this function is to generate

If that all makes sense, the key OpenCV call that I made was this: // The purpose of this function is to generate

If that all makes sense, the key OpenCV call that I made was this: // The purpose of this function is to generate

If I understand what you are trying to do, I have done something like this before. Basically the idea is to do an iterat

Relying on color: You can probably differentiate 15 markers based on color, but significantly more than 15 will be chal

Look into background modeling. There has been a lot of work on this subject. A "quick and dirty" approach would be to

I would also add that it is much easier to get calibration points into the far corners of the image if you use the Charu

I have done testing with the fisheye calibration vs standard calibration, and I found that the normal calibrateCamera fu

I have done testing with the fisheye calibration vs standard calibration, and I found that the normal calibrateCamera fu

My first guess would be that the intrinsic calibration is not very accurate (focal length in particular). How many imag

One more thing is that with high distortion you might run into issues where the calibration either doesn't detect the co

One more thing is that with high distortion you might run into issues where the calibration either doesn't detect the co

A few ideas: If your calibration target and camera can be held in a fixed position, take multiple images and compute an

I can't tell from your notation what you are doing, but if you multiplying the two vectors component by component, I don

I'm not sure what the differences are between estimateAffinePartial2D and estimateRigidTransform, but I use the former (

I'm not sure what the differences are between estimateAffinePartial2D and estimateRigidTransform, but I use the former (

Add an Aruco marker somewhere, and use that to determine orientation, or use a Charuco claibration target (instead of th

In my experience the focal length shouldn't change by that much from one calibration to the next (unless you are changin

For the 3 camera case you may want to look into using the trifocal tensor instead of two stereo calibrations. I'm not a

What looks wrong to you about the image? If it is the curved black boundary, that is normal because when you undistor

The intrinsic reprojection error is high, and the error from SolvePNP is absurdly high. I'd start by trying to address

I have had problems with the fisheye calibration with OpenCV 3.2. As I recall my issue was that it would run "successfu

I haven't had a chance to look at the code yet, but I did examine the image more closely. One thing that I noticed is t

Post your code.

YUYV format encodes two pixels in 4 bytes. The same 2 pixels as RGB use 6 bytes, so information is lost converting to YU

YUYV format encodes two pixels in 4 bytes. The same 2 pixels as RGB use 6 bytes, so information is lost converting to YU

@LBerger Have you had a chance to test with synthetic data? I'd be interested to know the results if you have any.

How do you compare the intrinsics and extrinsics of the calibrated vs synthetic camera models? It is just by looking at

If you are saving the image as a JPG or other lossy format, you could be seeing compression artifacts that result in R a

Using synthetic data with noise is a good idea. For this case I would add noise to synthetic input images (not image lo

Using synthetic data with noise is a good idea. For this case I would add noise to synthetic input images (not image lo

Using synthetic data with noise is a good idea. For this case I would add noise to synthetic input images (not image lo

The aruco corners aren't nearly as accurate as the chessboard corners, particularly in OpenCV 3.3.1. The chessboard cor

The aruco corners aren't nearly as accurate as the chessboard corners, particularly in OpenCV 3.3.1. The chessboard cor

I prefer these methods to Hough transform: link text I was able to get repeatable and robust results, but my input im

I prefer these methods to Hough transform: link text I was able to get repeatable and robust results, but my input im

The aruco corners aren't nearly as accurate as the chessboard corners, particularly in OpenCV 3.3.1. The chessboard cor

I have dealt with this type of situation by doing something like: For all of the lines: angle = (angle + 45) mod 180

I have dealt with this type of situation by doing something like: For all of the lines: angle = (angle + 45) mod 180 co

Yep...you are probably right...I totally missed the Anaconda part of the question...I was just focused on the assertion.

I'm not sure I understand what you are looking for, but maybe I do? First you need to figure out what your image center

What result are you looking for? An image with just the "fuzzy edge" pixels turned on, or a list of X,Y locations that

How consistent is the surface quality, and how consistent is the lighting? I might be inclined to try to isolate the th

see: Color Constancy Here is a paper (I haven't read it) that classifies road scenes. Maybe this method could be appl

see: link text Here is a paper (I haven't read it) that classifies road scenes. Maybe this method could be applied to