Hessian-Based Regularization for 3-D Microscopy Image Restoration

We investigate a non quadratic regularizer that is based on the Hessian operator for dealing with the restoration of 3-D images in a variational framework. We show that the regularizer under study is a valid extension of the total-variation (TV) functional, in the sense that it retains its favorable properties while following a similar underlying principle. We argue that the new functional is well suited for the restoration of 3-D biological images since it does not suffer from the well-known staircase effect of TV. Furthermore, we present an efficient 3-D algorithm for the minimization of the corresponding objective function. Finally, we validate the overall proposed regularization framework through image deblurring experiments on simulated and real biological data.


Published in:
Proceedings of the Ninth IEEE International Symposium on Biomedical Imaging: From Nano to Macro (ISBI'12), Barcelona, Kingdom of Spain, 1731–1734
Year:
2012
Publisher:
IEEE
Laboratories:




 Record created 2015-09-18, last modified 2018-09-13

External links:
Download fulltextURL
Download fulltextURL
Download fulltextURL
Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)