2016 Fiscal Year Final Research Report
Development of Medical Image Processing Technology Based on Sparse Coding and Hardware Implementation
Project/Area Number |
26420350
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Communication/Network engineering
|
Research Institution | Kanazawa University |
Principal Investigator |
|
Project Period (FY) |
2014-04-01 – 2017-03-31
|
Keywords | Sparse coding / Medical image processing / Noise reduction / Non-local means / Hardware design |
Outline of Final Research Achievements |
In this study, we developed a noise reduction method based on sparse coding for X-ray image as medical image processing algorithms, and we studied hardware implementation for the purpose of practical use. The major achievements were development of a dictionary design technique for sparse coding corresponded to Poisson noise and a fast atom selection method using initial inner product. These proposed method provided good noise reduction performance. On the other hand, we designed hardware of a fast processing algorithm using Cholesky decomposition, and evaluated the performance.
|
Free Research Field |
Image Processing
|