EAI Endorsed Transactions on Ambient Systems 17(16): e4

Research Article

Lorentzian Norm based Super-resolution Reconstruction of Brain MRI Image

Download18 downloads
  • @ARTICLE{10.4108/eai.19-12-2017.153485,
        author={Dongxing Bao and Xiaoming Li and Jin Li},
        title={Lorentzian Norm based Super-resolution Reconstruction of Brain MRI Image},
        journal={EAI Endorsed Transactions on Ambient Systems},
        volume={17},
        number={16},
        publisher={EAI},
        journal_a={AMSYS},
        year={2017},
        month={12},
        keywords={MRI image, super resolution image reconstruction, Lorentzian norm, regularization; iteration.},
        doi={10.4108/eai.19-12-2017.153485}
    }
    
  • Dongxing Bao
    Xiaoming Li
    Jin Li
    Year: 2017
    Lorentzian Norm based Super-resolution Reconstruction of Brain MRI Image
    AMSYS
    EAI
    DOI: 10.4108/eai.19-12-2017.153485
Dongxing Bao1,2, Xiaoming Li3, Jin Li1,*
  • 1: College of Automation, Harbin Engineering University, Harbin 150001, China
  • 2: School of Electronic Engineering, Heilongjiang University, Harbin 150080, China
  • 3: Dept. of Microelectronics Science and Technology, Harbin Institute of Technology, Harbin 150001, China
*Contact email: lijin@hrbeu.edu.cn

Abstract

Nowadays, SRR (super resolution image reconstruction) technology is a very effective method in improving spatial resolution of images and obtaining high-definition images. The SRR approach is an image late processing method that does not require any improvement in the hardware of the imaging system. In the SRR reconstruction model, it is the key point of the research to choose a proper cost function to achieve good reconstruction effect. In this paper, based on a lot of research, Lorenzian norm is employed as the error term, Tikhonov regularization is employed as the regularization term in the reconstruction model, and iteration method is employed in the process of SRR. In this way, the outliers and image edge preserving problems in SRR reconstruction process can be effectively solved and a good reconstruction effect can be achieved. A low resolution MRI brain image sequence with motion blur and several noises are used to test the SRR reconstruction algorithm in this paper and the reconstruction results of SRR reconstruction algorithm based on L2 norm are also be used for comparison and analysis. Results from experiments show that the SRR algorithm in this paper has better practicability and effectiveness.