Raymond's profile - activity

I was able to create already an eye detection application real time and static image using the OpenCV face detection sample. But sometimes it gives false detection and it can't detect faces in specified angle. I would like to hear your thoughts of what method might be better to use for eye detection of a moving (real time) and static images.

Using the opencv face detect sample I was able to detect faces in a live stream and detect face in an image by converting bitmap to mat and use the same approach as opencv face detect sample. I would like to ask, How can I do it in a video file like avi or mp4. I already loaded a video in a videoview but I don't know how to make the video file as an input so I can detect faces. Guidance, tutorial links, steps, samples is needed. Any will do. Thank you so much. Don't know how to read c++.

Is it possible to detect pupil using cascade classifier? If yes, will it be possible detect pupil in red and black? Can face detection be applied on it just like face detection can be applied to eye detection?

sorry for this, though I appreciate the help. I just wanted to clarify though that I googled it for a long time and didn't find the answer. Then when I use the keyword provided by LorenaGdL it appears. I will reflect in this action and won't do it again next time. Though I am glad that RolandC helped me out in my question. At first I thought that it was opencv line of code as I am new to opencv and not familiar with C and as I search I wasn't able to find it. Sincerely I apologize for this question and very thankful for the answer.

if (area >= 30 && std::abs(1 - ((double)rect.width / (double)rect.height)) <= 0.2 && std::abs(1 - (area / (CV_PI * std::pow(radius, 2)))) <= 0.2)

What I want to be translated is the std::abs and the CV_PI * std::pow

I've search the net for std::abs equivalent to java and failed to find any. Please help. Wasn't able to search for CV_PI * std::pow since I can't still find the std::abs

Does OpenCV support the comparison of video and image data set, returning some value (maybe a percentage) that indicates similarity? I'm trying to create an app that detects the eye glow and compare it in a data set of images of eye glows that I want to detect. I already detect the eye glow. Can I use it to compare to my data set of images of eye glows that I want to detect?