I am working on Face detection using Core Image. I m facing a problem when giving the points on face boundaries. Actually I ve to give the points on face boundaries and make those points in a movable state.

Please share your ideas with me.. Thanks in Advance...

• ios
• drawrect
• face-detection

I will suggest you four ways. I have not done them in IOS. I implemented them in java. For face boundary detection you can use following methods:

• Skin color Thresholding followed by chin curve estimate and shrinking and growing algorithm.
• Canny Edge Detection followed by four region divide and getting longest face edge map.
• Adaptive active contour model (Snake algorithm).
• Hair region detection followed by chin curve esimate and removing all the region except hair and chin. (Doesn't work for bald.)

If there are shadows in the face use Snake algorithm. If the faces in the images are clear then go for Skin Color Thresholding. Canny Edge Detection can give u very slow performance.

For further clarification read this : FACE REGION DETECTION