Font Size: a A A

Seal Identification Bank Notes

Posted on:2014-12-21Degree:MasterType:Thesis
Country:ChinaCandidate:Y PengFull Text:PDF
GTID:2268330425968365Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
With the rapid development of modern economy, the range of banking business is becoming more and more wider, also the volume of business that need to be dealt with is increasing, however, the seal has been widely used as the authentication of bank notes in the modern bank bills recognition system. The traditional seal recognition is often finished by artificial methods which have many disadvantages,such as slow speed, complex process, and low accuracy. In order to meet the needs of the development of the objective situation, improving work efficiency and realizing automatic identification of the notes is in urgent.Extracting and repairing of the seal is the basis of the seal recognition. When stamp a seal, due to the uneven force as well as the uneven color factors,the color of the seal will looks uneven,what’s more, the seal may be fuzzy and incomplete, plus bills itself has different gray,different color shading and words, all of this will make the extraction and repair of the seal face great difficulties. From the RGB color model, this paper proposes a new seal extraction algorithm that can extract a clear and complete seal image. If the seal has defect which is caused by the stamp, then we save it as a class of security features, do not fix it, if it is caused by the limitations of the image processing method, then we can use morphological methods and symmetric distribution algorithm to repair the seal.The match of the seal is the core of the seal recognition, and the match is good or bad will directly affect the result of seal recognition. With the continuous development of forgery techniques, many seal forgery technology has reached the level of real ones, plus there is a big difference between the real seal which make the seal recognition face greater difficulties. Therefore, this paper intends to improve the speed and accuracy of the detection by positioning the seal, adding a coarse discrimination step which set of three judgment conditions that corresponding to the real seal, the suspicious seal and the false seal respectively, for the suspicious seal, we need to judge it in the next seal recognition, for the real and false seal, we do not need the next seal recognition and give the result directly to improve the accuracy of seal detection. Now the common seal identification methods including the seal recognition based on invariant moment, the seal recognition based on circle-projection and the seal recognition based on the difference image, this thesis combines the difference image and the template matching methods to realize the seal distinguish.In modern society, studying the seal automatic recognition algorithm has high practical and theoretical significance. In this paper, the identification strategy can realize the automatic recognition of the seal, and the recognition speed is fast, meantime, the recognition rate is high. The algorithm has go through the matlab7.0simulation, and this detection has been applied to the bank bills recognition system.
Keywords/Search Tags:automatic seal identification, seal extraction, seal repair, distancedetection, similarity measure
PDF Full Text Request
Related items