Digital Image Morphing through Field Morphing
Download as PPTX, PDF2 likes1,508 views
This document describes a technique called "field morphing" for generating intermediate frames during digital image morphing. It uses line features specified in both source and target images, with correspondences established between lines. For each intermediate frame, a set of interpolated line features is generated. A warp function is computed based on displacements from pairs of corresponding line features, weighted by distance. Warping and color interpolation produces the morphed image. Pseudocode and a MATLAB implementation demonstrate the algorithm.
Digital Image Morphing through Field Morphing
- 1. DIGITAL IMAGE MORPHING THROUGH “FIELD MORPHING” Ankur Sakhuja
- 2. Example
- 3. Block Diagram for Algorithm Workflow Generating an intermediate frame Image IS Image I1 Fs → F warp Feature Spec Feature Fs Spec F Morph Color interpolate Image Image I2 Image IT FT → F warp Feature Feature Spec F Spec FT
- 4. Algorithm overview Basic motivation: specification of feature points as lines – more expressive and intuitive Line features specified in both source and target images and correspondence established For every intermediate position in morph sequence, a line feature set is generated by interpolating the two sets Between source and intermediate line feature sets:- Every pair of line features represents a coordinate transformation for a point from source to target image – results in some spatial displacement A weighted sum of displacements due to all line pairs gives net displacement of a point – the total warp function Warp both source and target images to get two intermediate images Color interpolate to obtain the morph image Repeat for every position in the sequence to obtain the morph sequence
- 5. The Math: Pixel transformation specified by single pair of line segments
- 6. The Math (contd): Pixel transformation specified by two pairs of line segments
- 7. The Math (contd): weighting factor for combining transformations of multiple line segment pairs b Weighting factor for combining displacements: length p Weight a dist
- 8. Algorithm Pseudocode • For each pixel X in the destination • DSUM = (0,0) • weightsum = 0 • For each line Pi Qi • calculate u,v based on Pi Qi • calculate X'i based on u,v and Pi'Qi' • calculate displacement Di = Xi' - Xi for this line • dist = shortest distance from X to Pi Qi • weight = (lengthp / (a + dist))b • DSUM += Di * weight • weightsum += weight • X' = X + DSUM / weightsum • destinationImage(X) = sourceImage(X')
- 9. MATLAB Implementation and Results
- 10. Reference 1. Beier, T. and Neely, S. 1992. Feature-based image metamorphosis. In Proceedings of the 19th Annual Conference on Computer Graphics and interactive Techniques