Design Of Banknote Image Acquisition And Processing System Based On SOC

The modern economy is developing rapidly,free trade between countries becomes more and more frequent.The currency detector is widely used in the field of banknote recognition and authentication.The image acquisition and processing system of banknote detector is the core module of banknote detector.Image acquisition hardware system,image data transmission and image correction are the key technologies of banknote image processing system.Through the research on image processing system of banknote detector,the main research work of this paper is listed as follows:This paper designs a single-processor multi-spectral image acquisition and processing system based on ZYNQ SoC processor.To solve the problem of insufficient for multispectral image data transmission in real time,this paper designs a method for data transmission based on high bandwith on-chip bus interface,using two dual port RAM for on-chip buffering,and using DMA to realized automatic transmission of image data,the performance of the high bandwidth and low latency have been achieved.This paper analyzes the quality of images acquired by cantact image sensors,a linear fitting algorithm is desgned to correct image which realized by hardwre,which reduced calculation time and improved real-time performance.In the view of the difference between different sensors,this paper designed the automatic exposure control based on PID algorithm,then,the image consistency of brightness has been improved.The experimental results show that the system can collect more type of spectral images of banknotes compared with traditional banknote detector.The hardware structure of data transmission on-chip can transmit image data in real time and quickly.The correction algorithm designed in this paper can improve the quality of image effectively,and is also has strong robustness of images quality collected by different sensors.