基于感知哈希和MOSSE算法的多国钞票类型快速识别技术研究

Research on Fast Recognition Technology of Multi-country Banknote Types Based on Perceptual Hash and MOSSE Algorithms

提出了用于现金自助设备的钞票类型快速识别技术。该技术通过综合利用图像感知哈希和MOSSE(Minimum Output Sum of Squared Error)算法实现,利用图像感知哈希快速计算机制,初步筛选出可能的钞票类型,再利用MOSSE相关滤波器最终确定钞票类型。

This paper proposes a fast banknote type recognition technology for cash self-service equipment.The technology is realized by comprehensively using image perceptual hash and MOSSE(Minimum Output Sum of Squared Error)algorithms.The possible banknote types are preliminarily screened out by using the fast image perceptual hash calculation mechanism,and the banknote types are finally determined by using MOSSE filters.