Multi-resolution network characteristic registration-based method for sorting face values and face directions of notes in sorter

Abstract

The invention relates to a multi-resolution network characteristic registration-based method for sorting face values and face directions of notes in a sorter, which comprises the following steps of: firstly, before performing image recognition and image measurement, balancing the luminance of different points on sampling images; secondly, performing Haar wavelet transform on note images, and extracting approximate picture contents of the note images; thirdly, extracting network characteristics; fourthly, analyzing the separability of the network characteristics on different note images; and finally, establishing a comprehensive sorter by using the network characteristics, wherein a structural risk minimization-based Gaussian mixture model is applied in the sorter. By the method, the real-time processing requirement of the note sorter can be met, and the requirement of processing 1,000 notes per minute can be met. Simultaneously, the defect of inconsistent note images in the same currency and face direction caused by abrasion and printing can be overcome effectively by adopting local characteristic registration so as to perform better recognition.

Description

Technical field: [0001] The invention relates to a face value-oriented classification method for banknotes, in particular to a face value-oriented classification method for sorting machine banknotes based on multi-resolution grid feature registration. Background technique: [0002] With the continuous prosperity of the market economy, the circulation of banknotes is increasing. Workloads for banking and retail workers are also getting bigger. But at present, the banknote sorting machines used by many domestic banks are imported from abroad and are expensive. The core technical basis of the sorting system is real-time banknote image processing and image recognition. There are few domestic banknote sorting machines, and their functions are very limited. It is difficult to meet the high-speed and real-time requirements, especially the sorting machines that can use image processing to identify banknotes have just started. Therefore, it is very necessary to develop a recogniti...

AI Technical Summary

Problems solved by technology

As the most effective banknote classification method, the neural network has been widely used in the banknote sorting system. For some banknotes with a high degree of similarity, the neural network can achieve excellent classification results, but because the neural network has a fatal shortcoming, It is easy to over-identify, that is, because the learning machine is too complex, in order to ensure high classification accuracy (small empirical risk), the VC dimension becomes very large, so the expected risk becomes very high

In the end, the learned samples have a high recognition rate, but the unlearned samples will be misrecognized or rejected.

At the same time, due to the lack of effective refusal means, it is easy to misidentify some similar printed matter as specific types of banknotes

Images