Fractal Techniques in Image Segmentation

Thursday, January 18, 2001 - 3:30pm - 4:30pm
Keller 3-180
Ning Lu (Advanced Digital Imaging Laboratories, Inc.)
The task of image segmentation, similar to the tasks of image compression and image enhancement, is used to be achieved by understanding the characteristics and geometry of images. The more knowledge - in either their frequency distributions or focused subject matters - we can apply, the better results can be achieved.

Extended works in fractal image compression and fractal image enhancement have proven that fractal technique is a unique powerful technique in digital image processing that can achieve very good results without explicitly enumerating knowledge from image science as most nonlinear algorithms found in engineering literature. Nevertheless, nonlinear algorithms are expensive for hardware implementation.

Here we present some further study in fractal image segmentation.