The Hardware Design And Implementation Of Banknote Multi-Spectral High-Speed Images Acquisition And Processing System

Banknotes are of great significance to a country’s financial security, its authentication has always been key topics in the field of anti-counterfeiting. Level-A banknote counter, banknote sorting machine and other intelligent financial machines, which has advantages on banknote image recognition and authentication, has been used in bank system gradually. Image acquisition and processing module is the key part of these intelligent financial machines, but always can’t meet the requirements upon banknote processing speed, stability, real-time property, banknote image recognition and authentication accurateness. So the research and implementing an efficient banknotes multi-spectral image acquisition and processing hardware platform is of great practical significance and application value.Aiming at the above problems, this thesis designed and implemented a hardware circuit module of banknote Multi-spectral image acquisition and processing based on Contact Image Sensor. this thesis has done the following work: 1. the thesis proposed a FPG A+DSP+AR M hardware solution by analyzing the existing high-speed image acquisition and processing technology, combined with the real situation of the banknote image acquisition and processing. 2. Based on this solution, the thesis selected key chips, analyzed feasibility of the solution and completed the circuit and PCB design. 3. Considering the existing solution of designing Network and USB interfaces occupies too much DSP processing time, this thesis used HPI interface of the DSP chip to fulfillment these interfaces, which enhances the work efficiency of the DSP chip greatly. 4. For some signals’ integrity problems the circuit system may exist, this thesis has done some simulation experiments and optimized the PCB design based on the simulation results, which greatly enhances system’s stability.The hardware platform described in this thesis has passed the authority apartment’s property test, and has also been tested in several intelligent financial machines. The results indicate that it has great adaptation ability and has great advantages on banknote processing speed, system’s stability and real-time property, banknote image recognition and authentication accurateness. compared with the corresponding module made by others.