Graver's profile - activity

Hello!

I need only to detect some squares in a image. The squares are found with canny(). The problem is that i need to binarize a HSV image. I tried to make the sum of the channels but now i have some gray in image.

I have a image in BGR color space and I convert it in HSV color space. I want to find a method that works on GPU like inRange works on CPU to binarize my image and obtain a threshold. Afer the conversion in HSV colorspace I make a split in three channels. After that I apply threshold() on every channel and finaly call function merge().

The problem is that in the last step i don't get the binary image. The image is colored.

Here is my code

gpu::cvtColor(myFrame_device, myFrameHSV_device, COLOR_BGR2HSV);
gpu::GpuMat channels_device[3];
gpu::GpuMat channels_device_dest[3];
gpu::split(myFrameHSV_device, channels_device);
//Mat myFrameEqualizedHSV(myFrameHSV_device);
//threshold HSV
gpu::threshold(channels_device[0], channels_device_dest[0], 0, 225, THRESH_BINARY);
gpu::threshold(channels_device[1], channels_device_dest[1], 75, 225, THRESH_BINARY);
gpu::threshold(channels_device[2], channels_device_dest[2], 0, 225, THRESH_BINARY);
gpu::merge(channels_device_dest,3, threshold_device);

I have a MFC project in Microsoft Visual Studio 2010. In my dialog I have a PictureControl. I generate a window with opencv and then i translate it to my PictureControl like this

namedWindow("INPUT", 1);
HWND hWnd = (HWND) cvGetWindowHandle("INPUT");
HWND hParent = ::GetParent(hWnd);
::SetParent(hWnd, GetDlgItem(IDC_BMP_LOAD)->m_hWnd);
::ShowWindow(hParent, SW_HIDE);

When I compile it for 32-bit arhitecture it works. When I compile for 64-bit it generate the namedwindow but doesn't hide and at the name I have a strange string like "%HO", different than "INPUT".

I tried to see if the window handler hWnd is ok and at unsued filed it gives me "CXX0030: Error: expression cannot be evaluated". In documention of OpenCV at the function cvGetWindowHandle() says that it returns a native HWND of the window refered by name in case of Win32 application.

Any help would be appreciated! Thanks.

Yes is simple. My problem is that I can't get the offset of the ROI because my ROI was created with RotatedRect. When i try to use locateROI it doesn't gave me the offset. After i get the offset i will make the rotation of points with the angle of RotatedRect "see the code".

Hello!

I have a little problem with som projection and geometry. I have an image where I detect a square. After the square detection, I crop the square from image. In the ROI I detect the point P(x,y) (see the image below).

My problem is that I know the coordinate of point P in the ROI, the coordinates of A,B,C,D, and rotation of ROI (RotatedRect::angle) but i want to get the coordinate of P in original image. Any advice could help.

Thanks!

EDIT :

For ROI crop I have this code

    vector< RotatedRect > rect(squares.size());

for (int i=0;i<squares.size();i++) // square.size() is the number of detected square(in some cases could be more squares in original image)
{
    rect[i] = minAreaRect(Mat(squares[i]));
    Mat M,rotated,cropped;
    float angle = rect[i].angle;
    Size rect_size = rect[i].size;

    if (rect[i].angle<-45)
    {
        angle += 90;
        swap(rect_size.width,rect_size.height);
    }

    M = getRotationMatrix2D(rect[i].center,angle,1.0);
    warpAffine(cameraFeed,rotated,M,cameraFeed.size(),INTER_CUBIC);
    getRectSubPix(rotated,rect_size,rect[i].center,cropped);
    cropped.copyTo(SatelliteClass[i].m_matROIcropped);
    SatelliteClass[i].m_vecRect = rect[i];
}

I have a little piece of code that try to convert a Mat to GpuMat and reverse. When I try to upload the Mat in GpuMat with "upload" function, it breaks. If I go next step in the debug the size of GpuMat is written, same as Mat but it doesn't upload the pixel data. Here is the code:

Mat initFrame;
Mat initconv;
Mat threshold;
imageCircle.copyTo(initFrame);
cvtColor(initFrame,initconv,CV_BGR2HSV);

gpu::GpuMat myFrame;
gpu::GpuMat myFrameLab;

myFrame.upload(initconv);

myFrame.download(threshold);
//gpu::cvtColor(myFrame,myFrameLab,CV_BGR2YCrCb);
imshow("INPUT",threshold);

I read that I will need Opencv built with GPU support. Here I don't know what to do. I saw somewhere that in OpenCV directories it exist /gpu/built/... I have 2.4.9 and 3.1 versions but I don't find in Opencv directory the /gpu/build/...

My problem was solved! Just a little adjustemnt at the code (for anybody who looks at this post). The index "i" need to go only to contours.size()/2, beacause if it goes after the nearest_circles_idx will be -1 and will be out of range of the array visited[] and will give an error of "Vector<bool> : iterator not dereferencable".

Thank you very much Guanta for your help!

Thanks for help!

Hello!

I wrote a code for detcting some circle, but i dont understand how the Hough Transform sort the output.

My program is runing like this

1. I have six circle that are positioned like this

1. i need to detect the cirlces and then make de half distance between the ccenters of circles two by two like this

• the blue dots are the half distances between circles two by two

Now the real example is something like this

For case 1 it works

But when i process a photo like this

Circles are sorted like this :

HERE IS MY CODE : http://pastebin.com/9WrsMtNQ