Sheng Liu's profile - activity

Thanks! I have tried the model in 2.4.13 and 3.0.0, and find little difference between the output of parvo and magno. So

Thanks! I have tried the model in 2.4.13 and 3.0.0, and find little difference between the output of parvo and magno. So

Thanks! I have tried the model in 2.4.13 and 3.0.0, and find little difference between the output of parvo and magno. So

Thanks! I have tried the model in 2.4.13 and 3.0.0, and find little difference between the output of parvo and magno. So

Thanks! I have tried the model in 2.4.13 and 3.0.0, and find little difference between the output of parvo and magno. So

Thanks! I have tried the model in 2.4.13 and 3.0.0, and find little difference between the output of parvo and magno. So

Which version should I choose? I need to use the retina model in OpenCV. And I find the model exist both in OpeCV2.x and

I'm sorry, but I cannot catch the meaning of that sentence. It seems like German.

Thanks a lot！

I intend to use multithreading to capture and process the video from my webcam, so I difine a global list to store frames. In 2.4 list<Mat>framelist;is available. However, the same code cannot be compiled in 3.0. How should I create list<Mat> in 3.0? Thanks!

My boss let me write a simple monitoring program which could record video only when people appear. But the equipment we have, outdated commputer and three simulation video cameras with a capture card, are both bought many years ago. I don't know how to use opencv(2.4.12) to play video from old capture card(connect PC via USB). All methods I tried are failed. My code is as follows:

CvCapture *capture1 = cvCreateCameraCapture(0);
    IplImage* img1;
        cvNamedWindow("a", CV_WINDOW_AUTOSIZE);
        while (1)
        {
            img1 = cvQueryFrame(capture1);
            if (!img1)break;            
            cvShowImage("a", img1); 
                }

The result of it is " ERROR:SampleCB() - buffer sizes do not match". Someone said if add " CV_CAP_DSHOW", " videoinput" mode woud be used. However, the code doesnot work well.

 CvCapture *capture1 = cvCreateCameraCapture(CV_CAP_DSHOW);
        IplImage* img1;
            cvNamedWindow("a", CV_WINDOW_AUTOSIZE);
            while (1)
            {
                img1 = cvQueryFrame(capture1);
                if (!img1)break;            
                cvShowImage("a", img1); 
    }

The result is still " ERROR:SampleCB() - buffer sizes do not match". How to solve the adaption problem? Please give me some help. Thanks!

Thanks! But I really need to know how to set that parameter.

HOG feature include gamma adjustment. Could you tell me how to set the parameter of gamma adjustment when I select HOG feature in opencv cascadeclassifier.

@ANirudh_Nandavar: I'm sorry to reply you so late. After pressing "Configure", you can find "BUILD_opencv_world" in "BUILD". Just now I didn't select BUILD_opencv_world but I compiled opencv3.0.0 successfully with cmake (3.3.0) and VS2013. Do you add extra mode? Maybe the version of opencv3 is not stable. Please try other version.

Thanks a lot！

Real Adaboost has a good result in fast rotation invariant multi-view face detection. RAB mode in opencv_traincascade is the same with the real adaboost proposed by Bo Wu?

Maybe you missed to add opencv_word300.

Please make sure your Vs and CUDA are all installed in default path in case of numerous errors.

I have a sailing boat model.The original image is here.

And this is the result of edge detection.

If I use StructuredEdgeDetection, can I just detect the edge of the hull(and ignore the sail). Need train new forest?

StructuredEdgeDetection and createRFFeatureGetter are too hard for me. I don't know how to use them to solve my own program. So could you give me some advice?

How to make the modelFile like "model.yml"?

Thanks to berak and 逸风, I find the LineSegmentDetector provided by this have a better result than LineSegmentDetector. The former output straight lines!