Ask Your Question

lasandcris's profile - activity

2014-11-15 16:13:06 -0600 answered a question Integration of SEEDS superpixels algorithm

Found it, thanks!

2014-11-15 06:26:10 -0600 commented answer Integration of SEEDS superpixels algorithm

hi, has this been implemented in version 3 (beta)? i saw in the online docs that there is now a set of functions for SuperpixelSEEDs but I can't find them in the build. thanks

2014-04-17 11:37:58 -0600 received badge  Scholar (source)
2014-04-17 11:35:24 -0600 commented answer GPU and Cuda

Thanks, I'm using 4.2. but I think it might be a graphics driver issue. I'm going to try to update. Its might effecting the runtime.

2014-04-17 07:13:43 -0600 asked a question GPU and Cuda


I'm trying to use gpu functions in my code but it seems something has gone wrong. I followed all the instructions out there that's related to setting up the gpu. Cmake recognized my cuda graphics card and I compiled all libs and bins including the new gpu dll which came to about 1.5gb. The strange thing is that when I use the getCudaEnabledDeviceCount() it returns a -1 value, which I'm not sure what it stands for. But I guess its not right. And of course when I use any other gpu related function it crashes the program. I'm using Visual Studio 2005, OpenCV 2.45, on Windows XP x64.

Does anyone know if there is a compatibility issue between Visual Studio 2005 and Cuda?

2014-04-17 05:40:03 -0600 answered a question Linker settings, gpu functionality removed or integrated with cpu?

Hi, I know its been a while since you had this issue but wondered if you managed to resolve this. I've built my own gpu lib and bin, and yes its ended up 1.5 gb only for the gpu dll. The problem is when I run the OpenCV samples it returns an error that the dll is not a valid windows image. The GetNumberOfCudaDevices function (or whatever is called) returns a '-1', not a 0. So I suppose it does recognize something, but no idea what the negative value means. Any suggestions?

2014-03-24 06:11:06 -0600 received badge  Self-Learner (source)
2014-03-24 05:54:24 -0600 answered a question RunTime Error at cornerSubPix call

In case anyone has similar issues, the problem was in the depth conversion of the image used in the cornerSubPix function. It seem to be fine when a single channel image is used.

2014-03-21 08:10:20 -0600 asked a question RunTime Error at cornerSubPix call

Hi guys,

I'm compiling a .dll (plugin) for a compositing software called Eyeon Fusion. I'm using the goodFeatureToTrack call and everything seems to be fine with generating the feature tracks etc.

The problem I'm having is when I introduce the cornerSubPix call. In all circumstances it has returned a RunTime Error and I've run out of the possibilities of what could go wrong.

I'm using Visual Studio 2008 as my compiler and I have also compiled the cmake file. Also, I'm using openCV 2.4.5.

Did anyone have any problems like this before and managed to solve it? Any suggestion would be appreciated.



This is the code snippet that causes the problem:

Size winSize = Size( 5, 5 );

Size zeroZone = Size( -1, -1 );

TermCriteria criteria = TermCriteria( CV_TERMCRIT_EPS + CV_TERMCRIT_ITER, 40, 0.001 );

cornerSubPix(EightBitSingle, cornersA, winSize, zeroZone, criteria);

2014-03-21 06:51:52 -0600 commented question lkdemo.cpp crashing upon cornerSubPix() and calcOpticalFlowPyrLK() call

Hi, have you managed to solve this problem. I'm having the same issue. Thanks lasandcris

2014-03-17 13:35:44 -0600 asked a question Input/OutputArrays

Hi, I'm just getting to know OpenCV and have a question on Input/OutputArrays. My plan is to integrate OpenCV into Eyeon Fusion. And to get myself familiar with the basics I'm using the cv::medianBlur(src, dst, ksize);

I'm using a vector of pixels from my image that I'd like to effect but it doesn't seem to be working. So, I'm not sure what type of vectors OpenCV is using to make this functional. This is the code snippet I'm testing:

for(int x = 0; x < inImage->Width; x++)             //loop through each pixel of the input image
    for(int y = 0; y < inImage->Height; y++)
        vector<float> vecR; // this is before these loops but couldn't post it like that for some reason

        p = sp;

        vecR.push_back((float(p.R)));              // assign each pixel to this vector

cv::medianBlur(vecR, vecR, 1);                  //using this vector for the blur(doesn't work)

int32 i = 0;

for(int x = 0; x < inImage->Width; x++)
    for(int y = 0; y < inImage->Height; y++)

        sp.GotoXY(x,y); dp.GotoXY(x,y);

        p = sp;
        p.R = vecR[i];   //write the new vector values back.
        dp = p;


So yes this is it and it doesn't seem to work. Also, if I change the kernel size to 3,5, 7 and so on, it crashes. But it could be because of the vector I'm using.

I would appreciate if someone could give me a heads up, it would help me a great deal on how OpenCV works.

Thanks, Laszlo