Hi everyone,

I've successfully upgraded to OpenCV 3.1 built and configured from source on my mac. When try to test the built libs I always get the following error (python for example)

(cv) sir:build sir$ python Python 2.7.10 (default, Jul 13 2015, 12:18:59) [GCC 4.2.1 Compatible Apple LLVM 6.0 (clang-600.0.57)] on darwin Type "help", "copyright", "credits" or "license" for more information.

import cv2 Traceback (most recent call last): File "<stdin>", line 1, in <module> ImportError: dlopen(/Users/sir/.virtualenvs/cv/lib/python2.7/site-packages/cv2.so, 2): Library not loaded: /usr/local/lib/libgflags.2.dylib Referenced from: /usr/local/lib/libopencv_sfm.3.1.dylib Reason: Incompatible library version: libopencv_sfm.3.1.dylib requires version 4.0.0 or later, but libgflags.2.dylib provides version 2.0.0

Now, since there is no version 4.0.0 of libgflags I wonder what's going on?

Anyone has an idea?

I'd really appreciate it, since I cannot compile my OpenCV-Based projects anymore atm disaster help :)

I'm trying to compile a project that I have written on Mac using QT and OpenCV3 on Windows.

I'm failing in the Link step - LNK2038 : mismatch detected for 'RuntimeLibrary' MTd_StaticDebug != MDd_DynamicDebug

Both BOOST and QT have issues when you do not use "dynamic C run time library " if I understand the respective documentations correct.

I can start compiling OpenCV on my Windows machine now, but last time I did that on mac i needed half a day to configure it correctly...

Is there a way I can get OpenCV precompiled with /MDd ? I'm sure a lot of people use OpenCV with BOOST Unit testing and QT, so everyone would be lost with the supplied download, or am I missing a point here?

Sincerely,

Wolfram

There are a lot of ways to do that...

One way it can be done is like at http://nghiaho.com/uploads/code/videostab.cpp

(frame processing snippet)

cap >> cur;

if(cur.data == NULL) {
    break;
}

cvtColor(cur, cur_grey, COLOR_BGR2GRAY);

// vector from prev to cur
vector <Point2f> prev_corner, cur_corner;
vector <Point2f> prev_corner2, cur_corner2;
vector <uchar> status;
vector <float> err;

goodFeaturesToTrack(prev_grey, prev_corner, 200, 0.01, 30);
calcOpticalFlowPyrLK(prev_grey, cur_grey, prev_corner, cur_corner, status, err);

// weed out bad matches
for(size_t i=0; i < status.size(); i++) {
    if(status[i]) {
        prev_corner2.push_back(prev_corner[i]);
        cur_corner2.push_back(cur_corner[i]);
    }
}

// translation + rotation only
Mat T = estimateRigidTransform(prev_corner2, cur_corner2, false); // false = rigid transform, no scaling/shearing

// in rare cases no transform is found. We'll just use the last known good transform.
if(T.data == NULL) {
    last_T.copyTo(T);
}

T.copyTo(last_T);

// decompose T
double dx = T.at<double>(0,2);
double dy = T.at<double>(1,2);
double da = atan2(T.at<double>(1,0), T.at<double>(0,0));

prev_to_cur_transform.push_back(TransformParam(dx, dy, da));

out_transform << k << " " << dx << " " << dy << " " << da << endl;

cur.copyTo(prev);
cur_grey.copyTo(prev_grey);

cout << "Frame: " << k << "/" << max_frames << " - good optical flow: " << prev_corner2.size() << endl;
k++;


(snippet)

Another way that is very robust is to take a few e.g. 64x64 gray blocks of the image at fixed positions and correlate them with the following frame. The position of the maximum value of the resulting matrix corresponds to the translation.

You might want to apply lens distortion before if you want to have real nice results and you're "zoomed out".

W.

The method

CvKNearest::get_sample_count ( Manual : Returns the total number of train samples. C++: int CvKNearest::get_sample_count() const )

seems to be a victim of the latest refactoring ... how to I get the train sample count now?

Thanks a lot for the quick reply. The method

CvKNearest::get_sample_count ( Manual : Returns the total number of train samples. C++: int CvKNearest::get_sample_count() const )

seems to be missing as well... what happend to it / how to I get the train sample count now?

Thanks again

Thanks a lot for the quick reply. The method

CvKNearest::get_sample_count ( Manual : Returns the total number of train samples. C++: int CvKNearest::get_sample_count() const )

seems to be missing as well... what happend to it / how to I get the train sample count now?

Thanks again

Thanks a lot for the quick reply. The method

CvKNearest::get_sample_count ( Manual : Returns the total number of train samples. C++: int CvKNearest::get_sample_count() const )

seems to be missing as well... what happend to it / how to I get the train sample count now?

Thanks again

Thanks a lot for the quick reply. The method

CvKNearest::get_sample_count ( Manual : Returns the total number of train samples. C++: int CvKNearest::get_sample_count() const )

seems to be missing as well... what happend to it / how to I get the train sample count now?

Thanks again

Thanks a lot for the quick reply. The method

CvKNearest::get_sample_count ( Manual : Returns the total number of train samples. C++: int CvKNearest::get_sample_count() const )

seems to be missing as well... what happend to it / how to I get the train sample count now?

Thanks again

Hi Ninja,

try to initialize named windows first

(...)

cv::namedWindow("RedCamera"); cv::namedWindow("BlackCamera");

for (;;){

(...)

cv::imshow("RedCamera", cameraFrame1); cv::imshow("BlackCamera", cameraFrame2);

(...)

(see also:)

http://stackoverflow.com/questions/19565262/opencv-imshow-does-not-work

Hi,

I have a perfectly working program written in c++ under Mac OS 10.9 (LLVM) with QT5 and OpenCV that i pulled some weeks ago from the development branch. It uses cv::KNearest.

I'm trying to compile this application under Windows using OpenCV 3.0.0 Alpha, QT5 latest version, Visual Studio '13 but don't succeed.

I get

cv::KNearest : Is not a member of 'cv' (?)

Looking into the <ml.hpp> class, I figured out to try to use cv::ml::KNearest instead, but now I get

C2259 cv::ml::KNearest : cannot instantiate abstract class

Can somebody help me with this? Has there been a recent change in the KNearest / ML code, ...?

Sincerely,

Wolfram