2017-04-10 11:36:59 -0600 | commented answer | OpenCV 3.2 problem with displaying properly text on live video Yeah it worked! Thanks! I changed font to HERSHEY_PLAIN and played with size. Now it's fine. |
2017-04-07 08:21:34 -0600 | commented answer | Printing distance vector for all faces in OpenCV Python FaceRecognizer class |
2017-04-07 08:19:26 -0600 | asked a question | OpenCV 3.2 problem with displaying properly text on live video I'm testing face recognition on Raspberry Pi 3 and since my first run of OpenCV 3.2 I've got some strange artifacts in displayed screen in places where text is shown (label with confidence for recognized person and date time). Earlier when I was using OpenCV 2.4 everything was fine. Any idea why is happening like this? Code: https://gist.github.com/anonymous/5c5... Photo: http://imgur.com/a/XOf41 |
2017-04-07 07:55:46 -0600 | received badge | ● Supporter (source) |
2017-04-07 07:55:40 -0600 | received badge | ● Scholar (source) |
2017-04-07 07:42:56 -0600 | commented answer | Printing distance vector for all faces in OpenCV Python FaceRecognizer class Yees, its's working! Thank you so much for your help, really appreciate it. Now I need to figure out how to plot something from this output. Maybe I can ask for another thing, since my first run OpenCV 3.2 on RPi3 I've got some artifacts in displayed screen in places where text is shown (label with confidence for recognized person and date time). Earlier when I was using OpenCV 2.4 everything was fine. Any idea why is happening like this? Photo below: [http://imgur.com/a/XOf41] |
2017-04-07 05:15:54 -0600 | commented answer | Printing distance vector for all faces in OpenCV Python FaceRecognizer class StandardCollector not found, the same with PredictCollector :/. I checked in docs both of them: http://docs.opencv.org/3.2.0/d4/d8d/classcv_1_1face_1_1StandardCollector.html#abd1bca19318b57f976677faf31a8cd03 (http://docs.opencv.org/3.2.0/d4/d8d/c...) http://docs.opencv.org/3.2.0/da/d6a/classcv_1_1face_1_1PredictCollector.html (http://docs.opencv.org/3.2.0/da/d6a/c...) I'm calling it with:
|
2017-04-06 14:07:46 -0600 | commented answer | Printing distance vector for all faces in OpenCV Python FaceRecognizer class Thanks, just compiled OpenCV 3.2 on my Raspberry Pi 3. Any idea how can I call this function "predict overload" in Python? Was looking into OpenCV 3.2.0 documentation, but nothing found :(. Is it some kind of array of results on output with all distance coefficients, that I could plot later? |
2017-04-06 05:20:28 -0600 | asked a question | Printing distance vector for all faces in OpenCV Python FaceRecognizer class I'm working with OpenCV 2.4.9.1, Python 2.7.9 and using FaceRecognizer class for my real-time face recognition university project. Need to do some research about existing algorithms (Eigenface, Fisherface, LBPH) and plot some data. Normally each algorithm is checking in certain way similarity of faces from database to that real one and append label with lowest coefficient. Is there any way to get predictions of all faces in database to one face that is in front of the camera, so I could draw some affilation function for each person or some face-distance graph?? It would be quite helpful in my analysis part of work. I saw that people can do that probably in C++ or when using SVM but no examples how to do it in python with existing opencv algorithms. |