Research On Coin Recognition

With the fast-paced development of the human society and coin circulation automation equipment unattended, all-day service to more and more consumers of all ages, are widely used at present in many public places. Automated coin circulation in the use of equipment use of counterfeit money posing as real currency is not uncommon, so coin recognition inspection device is also came into being. Coin identification detection devices currently in use is based on principle, circuit design, software and mechanical structure of the algorithm there is a technical defect, there is often costly and recognition accuracy is not high.This article elaborates a coin recognition with high accuracy, low cost detector design and its implementation of programme. Through analysis of Eddy current sensor principles, introduced the use of dual-channel Eddy current sensor to detect the parameters such as material, the thickness of the coin, coin recognition improved detection rate effect. At the same time, initial speed to overcome coin, coins in the coin in the impact on measurement of the beat phenomenon, using moving average filter algorithm for processing the collected data.To distinguish genuine coins. Coin identification devices with learning function, not only can you learn to recognize the coins issued by the Central Bank also can learn to recognize your special coin or token. The device features coin or token parameter values are saved in EEPROM memory, use this as a basis for identifying the authenticity of coins. The device also has a live update feature, when the Central Bank issued a new version of the coin or the Manager to increase when a new version of the token, just start learning all over again, would feature new currencies of characteristic parameter values in EEPROM memory, thereby increasing the recognition of new currencies. The device also has warning function, can sound alarm notes.Coin identification device of the core technical issues during the project’s research results have been successfully resolved.This project research content in the main innovation points: 1) swirl sensor used double frequency work way, improve coins identification device correctly identification rate;2) used speed weighted algorithm, elimination cast coin speed on average frequency value of effect;3) can online system upgrade, increased on SGD species of identification features;on coins recognition device repeatedly test, proved cast coin Manager system performance good, and stability reliable, on really coin and counterfeit money of identification rate respectively for98.9%and98.3%.Coin recognition through the design and commissioning of the device, can be summed up:characteristics such as texture, pattern coins, coin quality and exact dimensions can be fetched by Eddy current sensor. To visit the coin features, coin recognition accuracy substantially greater.