Neural concurrent subsampling and interpolation for images

作     者：Kim, JO Choi, BT Ko, SJ 

作者机构：Korea Univ Dept Elect Engn Seoul 136701 South Korea 

出 版 物：《ELECTRONICS LETTERS》 (电子学快报)

年 卷 期：1999年第35卷第22期

页      面：1933-1934页

核心收录：

学科分类：0808[工学-电气工程] 0809[工学-电子科学与技术（可授工学、理学学位）] 08[工学] 

主　　题：.high speed original image interpolation techniques FNN proposed technique interpolation algorithm feedforward neural network good image increased performance conventional methods image subsampling LPFS subsampling process subsamplcd Proposed subsampling 

摘      要：A concurrent subsampling and interpolation algorithm for images is presented which uses a feedforward neural network (FNN), providing the advantages of good image reproduction quality and high speed processing capability. Experimental results show that the proposed technique exhibits an increased performance over conventional methods such as linear- and cubic-based interpolation.