| Acronym: SCODE |
| Main Objective: The main objective of this work is to investigate and develop an usable version of a MMP based scanned documents encoder, that is able to improve on the encoding performance of current state-of-the-art compound image encoders, like DjVu. In order to achieve this, some important tasks must be achieved: - investigate and develop efficient methods for layer-based or block-based image segmentation, for optimisation of the overall MMP-based encoder�s performance; - optimise the dictionary adaptation process in order to increase the coding performance of MMP for the images corresponding to the segmented layers or block types; - efficiently implement a computational version of the method and investigate relevant functional optimizations, in order to reduce the computational complexity of the algorithm without compromising its coding efficiency; - perform a set of experimental tests in order to compare the encoding performance of the new method with the one of other state-of-the-art methods. |
| Reference: PTDC/EEA-TEL/66462/2006 |
| Funding: FCT/PTDC |
| Start Date: 01-10-2007 |
| End Date: 01-10-2010 |
| Team: Sérgio Manuel Maciel de Faria, Nuno Miguel Morais Rodrigues, Vitor Manuel Mendes da Silva |
| Groups: Multimedia Signal Processing – Lr, Multimedia Signal Processing – Co |
| Partners: |
| Local Coordinator: Sérgio Manuel Maciel de Faria |
| Links: http://www.co.it.pt/~scode/ |
|
|
|
Associated Publications
|
