HSV's profile - activity

I am also interested in motion capture using opencv. but have not found any prebuild modules. I think we need to start

thank you.

The large dots in the center are used to calculate the homography. What is good way to find only the large dots in the

not all dots are detected in each image, like the example, there are 3 or 4 rows of dots at the top of the board that are not being detected.

I found this implementation of osoftposit, https://github.com/hanshuo/softposit there are matlab fortran source files. I haven't had any luck getting those files to work, if anyone else does please post process! thanks.

Hi all,

I know the x, y, z values of all the points on this calibration board very precisely. I want to calculate a homography between the found pixel locations of the large dots, and their corresponding x, y, values. Then I can map the image, so the dots are in nicely arranged rows and columns, so then I can determine what row/column each dot is on.

Does anyone have any tricks / tips on how to determine the correspondence between the found large dot pixel u,v location and their known x,y locations?

Thank you!

the best answer to the question I got was from Francesco Callari over on stackoverflow best answer yet

I have used opencv to calibrate a stereo camera pair.

Using the determined Rotation matrix and Translation vector, I would like to be able to calculate the pose of the second(R) camera, given the pose of the first(L) camera.

I have a scene where I use solvePnP to calculate the pose of the Left camera, rvec and tvec. How can I determine the rvec and tvec of the Right camera?

Thanks

Hi all, newbie question here, I'm trying to use a dc1394 (firewire) camera with the python interface on Windows 7, but can't seem to get cv2.VideoCapture to output anything, i.e. doesn't find attached camera. I've tried the cv2.getBuildInformation() function, but don't see any reference to the dc1394 lib. Any suggestions would be great. Thanks. JB

For our particular job I've been developing some image segmentation routines using opencv cv2, python bindings and USB webcams. I'd like to keep my current work (cv2 python code) and integrate my new Point Grey Flea2 firewire cameras. I'm using Windows 7. I have yet to find an answer that actually works.

libcamiface, couldn't generate exe file pydc1394, fails with 'dc1394_new not found', but the cmu 1394camera demo program works.

thanks for the help. jvb

I was wondering if anyone has implemented the softPosit routine in c++ or python, and if so would share with me what they’ve done. I’m using windows 7 and python bindings, but any c++ code would be helpful.

the softposit routine does "for determining the pose of a 3D object from a single 2D image when correspondences between object points and image points are not known".

I have exactly that problem. I know the 3d locations of some targets. I then can use only 1 image of the object. From that image I find the targets, or most of them, sometimes targets are blocked from view. I then need a way to automatically calculate the pose. I can't use the Posit in opencv, because the 3d/2d lists must be same size, and in the same order. I don’t know the order without having to do some sort of manual picking. That is not acceptable any longer for my problem. Hence, softPosit, it can deal with 1st the correspondence problem, then the posit problem.
Thanks John

ref: http://www.umiacs.umd.edu/~ramani/pubs/softposit.pdf