Banknote recognition based on optimization of discriminative regions by genetic algorithm with one-dimensional visible-light line sensor

作者：

Highlights：

• Using similarity map, we build a selection priority map of sub-sampled banknote image.

• We optimize masks by using genetic algorithm (GA) based on selection priority map.

• Our method considers both distinguishability and optimality of selected image region.

• Our method shows the high recognition accuracy with 4 banknote databases.

摘要

•Using similarity map, we build a selection priority map of sub-sampled banknote image.•We optimize masks by using genetic algorithm (GA) based on selection priority map.•Our method considers both distinguishability and optimality of selected image region.•Our method shows the high recognition accuracy with 4 banknote databases.

论文关键词：Banknote recognition,One-dimensional visible light sensor,Genetic algorithm,Optimal discriminative region,Kinds of banknote databases

论文评审过程：Received 29 October 2016, Revised 19 June 2017, Accepted 25 June 2017, Available online 27 June 2017, Version of Record 5 July 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.06.027