Face Features Detection System - with OpenCV
Why Stack Overflow recommends it
Stack Overflow users link it as a concrete example of what OpenCV-based facial feature detection can achieve when answering or asking about face detail and landmark recognition.
Recommended by 2 people across 2 different months, 2011 to 2013.
What it is
A 43-second 2008 demo of a real-time face features detection system built with OpenCV running on a standard webcam. It shows detection of head position, eyes, eyelids, brows, nose, mouth, lips, teeth and nostrils, and can track more than one face at once. The author says the code will not be shared.
What people said
Recommends foundry · 2013-01 · Stack Overflow · on “How to Find position of Ears, cheek and Head positon for IOS using CIFaceFeature”
CIFaceFeature is derived from and inspired by academic research work that generally gets poured into openCV [LINK] . So to go beyond CIFaceFeature, you will need to get to grips with the openCV libraries. OpenCV.org provides a C++ framework that runs on most platforms, including iOS. In particular you will want to look at object detection [LINK] , the cascade classifier [LINK] and haartraining [LINK] . Judging by your comments you would like a simpler solution. I suppose that you could invent a generic 'face' type, idealised from everyone to a single example. Then calculate the relationships - from all angles?. That's as I have suggested, substantially simplifying the need for training. But
Recommends Leroy Meijer · 2013-03 · Stack Overflow · on “Recognizing a face in a set of images”
Take a look at OpenCV for C++ [LINK] I have worked with, and realised a application that could recognize playing-cards based on pre-loaded templates of those cards. But OpenCV is also excellent for recognizing (so templating) faces. [this video] Maby this is a good starting point?
Cites manneorama · 2011-01 · Stack Overflow · on “Looking For Realtime Facial Feature Detection/Extraction Library”
I'll let someone more experienced give a proper answer but in the meantime I know that OpenCV has support for detection of facial features. Here is the documentation: [LINK] , and here is a demonstration: [this video] Is that what you're looking for?
Cites ffriend · 2012-01 · Stack Overflow · on “Recognition of face details as a set of points, not just rectangles”
I'm doing research in the field of emotion recognition. For this purpose I need to catch and classify particular face details like eyes, nose, mouth, etc. Standard OpenCV function for this is detectMultiScale() , but its disadvantage is that it returns list of rectangles ( video [this video] ) while I'm mostly interested in particular key points - corners of mouth, upper and lower points, edges, etc ( video [LINK] ). So, how do they do it? OpenCV is ideal, but other solutions are ok too.