Euro coin nationality recognizing method based on local binary pattern

Abstract

The invention discloses a euro coin nationality recognizing method based on a local binary pattern. The euro coin nationality recognizing method based on the local binary pattern comprises the following steps: step one, acquiring gray level image of a coin, extracting a to-be-detected target region from the acquired coin image, and carrying out normalization processing on the size; step two, carrying out annular space disintegration on the target region; step three, extracting a unified rotating unchanged local binary pattern of each annular region, counting histogram distribution features of each annular region, and then assembling the annular regions according to the sequence from an inner ring to an outer ring to obtain fine coin image description features; and step four, detecting and recognizing euro coins with different currency values by respectively designing a plurality of classifiers through a support vector machine according to the different currency values. The area is used as a feature to determine the currency values of the euro coins, the euro coins are recognized by the different classifiers selectively according to the currency values, then a euro coin nationality recognizing problem is disintegrated, and the difficulty of the problem is reduced to guarantee accuracy of follow-up nationality recognition.

Description

technical field [0001] The invention relates to a country identification method for euro coins, in particular to an image-based country identification method for euro coins. Background technique [0002] Among the 28 member states of the European Union, 17 countries use the euro, of which 12 countries including France, Germany, Spain, Portugal, the Netherlands, Ireland, Finland, Greece, Italy, Luxembourg, Belgium, and Austria have a large circulation. Slovenia, Cyprus, Malta, Slovakia, and Estonia have a small coin market circulation. There are 8 types of euro coins in circulation on the market, namely 2 euros, 1 euro, 50 euro cents, 20 euro cents, 10 euro cents, 5 euro cents, 2 euro cents, 1 euro cent, of which 5 euro cents, 2 euro cents Cents, 1 euro cent, 3 kinds of currencies are less in circulation. All countries have the same obverse pattern on euro coins, while the reverse pattern is designed by each country. Therefore, the national identification of euro coins can...

AI Technical Summary

Problems solved by technology

[0006] The recognition of euro coin images has rarely been reported in previous literature, and the performance of related methods lacks comparative basis and benchmarks. According to market research, the designed algorithm should make the recognition error rate less than 0.5%. How to solve the above technical problems needs to be explored and studied The problem

Images