RAJAHMUNDRY A.P.

INDIA

Image Processing
--emphasis on face morphing BY Department Of Ch.V.V.D. Prasad K.Chandra Suresh III/IV B.Tech

Information Technology

GODAVARI INSTITUTE OF ENGINEERING & TECHNOLOGY

Introduction

Morphing is an interpolation technique used to create a series of intermediate objects from two objects. The face - morphing algorithm automatically extracts feature points on the face and morphing is performed. This algorithm is proposed by Mr. M.Biesel within Bayesian framework to do automatic face morphing.

Outline of the Algorithm

Pre Processing
removing the noisy backgrounds
clipping to get a proper facial image, and scaling the image to a reasonable size.

Feature Finding
Here we find 4 major feature points, namely the two eyes, and the two end-points of the mouth. Eye - finding

Feature Finding

We first compute the complexity map we multiply the complexity map by a weighting function. we find the three highest peaks in the weighted complexity map . Finally, the similarity is measured in the correlation-coefficient sense

Image Partitioning
Feature finder can gives the positions of the eyes and the end points of the mouth, so we get 4 feature points. We generate 6 more feature points around the face edge.
__________

Contd.

Image Partitioning
Based on these 10 feature points, 16 non-overlapping triangular or quadrangular regions developed

As the feature points are at different positions, we do warping between images.

Image1 Image2

Coordinate Transformations
Affine Transformation
A
q

C B

Coordinate Transformations
Bilinear Transformation

Cross-Dissolving
The last phase of face - morphing Cross-dissolving is described by the following equation,

where A,B are the pair of images, and C is the morphing result.

Example on typical morphing process.

Pre - Processing

__________ __________

Coordinate transformations

Cross-dissolve

Morphing between faces of different people

- human and animal (lion)

- man and man

- man and woman

Conclusion

We believe that feature extraction is the key technique for automatic face morphing algorithms. More and more feature points give the better result Therefore, in this presentation we emphasized on eye-finder. This algorithm, is successful in eye detection rate by 84%.

References

Martin Bichsel, "Automatic Interpolation and Recognition of Face Images by Morphing. Jonas Gomes, "Warping and morphing of graphical objects", Morgan Kaufmann Publishers (1999).