Neural Evolution Model for Gray Level Image Restoration
This chapter introduces a novel image restoration method by using modified evolution strategy (MES). We first review the background of image restoration and some useful concepts associated with it. A MSE image restoration model is made in Section 4.2. Three main improvements that consist of coordinate descending mutation, “survival of the fittest” selection rule and hybrid evolution strategies are presented in Section 4.3. Neural evolution algorithm is introduced in Section 4.4. Theoretical analysis and experimental results in Section 4.5 demonstrate the efficiency of the proposed method. In Section 4.6, we draw a conclusion.
Unable to display preview. Download preview PDF.
- D.B. FogelEvolutionary Computation: Toward a New Philosophy of Machine IntelligenceIEEE Press, Piscataway, 1995.Google Scholar
- W.K. PrattDigital Image ProcessingJohn Wiley & Sons, 2nd ed., New York, 1973.Google Scholar
- R. Chellappa (ed.)Digital Image Processing IEEE Computer Society Press1992.Google Scholar
- R.C. GonzalezDigital Image ProcessingAddison-Wesley,1992.Google Scholar
- J.C. RussThe Image Processing HandbookRoca Raton: CRC Press,1995.Google Scholar
- D. ZhangNeural Network System Design MethodologyTsinghua University Press, 1996.Google Scholar
- G.A. Carpenter and S. GrossberNeural Networks for Vision and Image ProcessingThe MIT Press, Cambridge, 1992.Google Scholar