mohamedadaly / caltech-lane-detection Goto Github PK

I am making a Spray Painting Robot and want this tech to implement in my Robot. Please guide me how to use this and implement.
The documentation needs information about implementing it.
I would love to contribute its documentation if I am explained and helped to understand the code better.

Email me at: [email protected]

Hello I am so interested with your research. Many thanks to provide the code for us. In measuring performance of this code, I download the caltech line dataset. However is there any label/ground truth for the dataset?
Thank you

Thank you very much for your open source project, thank you. When I saw the mcvGetVanishingPoint function, I was very puzzled about the calculation of the pitch matrix, Ithink it should beFLOAT_MAT_ELEM_TYPE tpitchp[] = {1, 0, 0,
0, cos(cameraInfo->pitch), -sin(cameraInfo->pitch),
0, sin(cameraInfo->pitch), cos(cameraInfo->pitch)};
or was there something wrong with my idea?

Hi @mohamedadaly,
Thanks for providing the source code.
I processed the code. It processed for few images and threw error after that. Following is the error:

OpenCV Error: Assertion failed (w.type() == type && (w.size() == cv::Size(nm,1) || w.size() == cv::Size(1, nm) || w.size() == cv::Size(nm, nm) || w.size() == cv::Size(n, m))) in cvSVD, file /build/buildd/opencv-2.4.8+dfsg1/modules/core/src/lapack.cpp, line 1746
terminate called after throwing an instance of 'cv::Exception'
  what():  /build/buildd/opencv-2.4.8+dfsg1/modules/core/src/lapack.cpp:1746: error: (-215) w.type() == type && (w.size() == cv::Size(nm,1) || w.size() == cv::Size(1, nm) || w.size() == cv::Size(nm, nm) || w.size() == cv::Size(n, m)) in function cvSVD

I am attaching the snapshot of terminal also.

Hi @mohamedadaly and thank you for you code! It is really interesting, but I'm implementing the code in order to detect the road lanes and I found that your configuration files are all related to the stop markers. Do you have any "config" file to the road lane detection?

Thank you in advanced.