| Line 13: | Line 13: | ||
In my above experiment i have added two images.The images were first cropped to have same size and then added and gamma corrected .In the above pictures,the morphed images are not the correct representations of image morphing. Morphing of images is based on pattern recognition and co-ordinate transformation(Bilinear transformation). | In my above experiment i have added two images.The images were first cropped to have same size and then added and gamma corrected .In the above pictures,the morphed images are not the correct representations of image morphing. Morphing of images is based on pattern recognition and co-ordinate transformation(Bilinear transformation). | ||
| − | ''' | + | '''So how does morphing takes place?''' |
---- | ---- | ||
1. An algorithm is run to extract the face features. | 1. An algorithm is run to extract the face features. | ||
Revision as of 02:14, 7 December 2009
MORPHING OF IMAGES
Morphing of images deals with transition from one image into an another image.It looks like an animation.Photo morphing is used to see what happens when two images are mixed.One practical use of morphing is see predict the child's face by morphing the parent's photos.
In my above experiment i have added two images.The images were first cropped to have same size and then added and gamma corrected .In the above pictures,the morphed images are not the correct representations of image morphing. Morphing of images is based on pattern recognition and co-ordinate transformation(Bilinear transformation).
So how does morphing takes place?
1. An algorithm is run to extract the face features.
2.Eyes and mouth are the main facial features.Eyes are traced by a complex analysis.This analysis is based on summing the intensity difference between adjacent pixels and then multiplied by a probability.The pdf is based on the probability of finding the eye in the face.
3.Then mouth is traced and the two ends of the mouth are found.
4.So four points on the face are found.Then some points on the facial boundary are found by edge detection.
5.Finally with all the 10 points ,the entire face is divided into triangles and quadrilaterals.
6.Following this with the help of co-ordinate transformation ,one-to-one transformation is done on two images used for morphing.
7.Finally we get the morphed image.
