Image fine hollow image-text identification method and device

Abstract

The invention discloses an image fine hollow image-text identification method and device. The method comprises the steps of obtaining an image of a banknote, and extracting a small image comprising afine hollow image-text region according to the coordinate information of the region where the fine hollow image-text is located; positioning a safety line region from the small image containing the fine hollow image-text region; and in the safety line region, calculating a plurality of columns according to equal-difference sampling, searching the distribution condition of the gray level jump points from top to bottom for each column of sampled pixel points so as to obtain the distribution condition of the gray level jump points of each column in the safety line region, and judging the authenticity of the banknote based on the distribution condition of the gray level jump points of the real banknote. According to the invention, the calculation is focused on the target fine hollow image-textinformation; the operation process is free of complex mathematical calculation and high in time efficiency; and in addition, extraction of the foreground hollow image-text does not depend on a binarization fixed threshold value, and image gradient information is sequentially calculated through sliding sequence units, so that fine hollow image-text information can be effectively extracted throughmathematical description, and robustness is higher.

Description

technical field [0001] The invention relates to the field of identification of banknote anti-counterfeiting points in image processing applications, in particular to a method and device for identifying fine hollowed-out images. Background technique [0002] In the circulation of banknotes, criminals often create counterfeit money or maliciously damage banknotes, disrupting the security of the circulation of the financial currency market. extension etc. These banknote detection devices, especially those used in banks and other financial institutions, are generally equipped with CIS image sensors, which can collect image information of banknotes. Banknotes under different spectra will have different image feature information. Based on this, banknotes can be detected Image analysis of the anti-counterfeiting points of the image. [0003] Among the common banknote anti-counterfeiting methods, in order to enhance the anti-counterfeiting features, anti-counterfeiting features of...

AI Technical Summary

Problems solved by technology

[0005] Combined with the actual use effect, the above method has the following disadvantages 1. The algorithm results are heavily dependent on the binarization segmentation effect, the difference in the imaging effect of different image sensors, the change of the foreground and background contrast of image calibration, the difference between the old and new target banknotes, etc., the threshold value calculated by binarization It is difficult to segment with the best effect, resulting in differences in the calculation of connected domain information, and the consistency of the calculation results of the fine hollowed-out graphic information of real coins is unstable; 2. The algorithm takes too long, the entire algorithm takes more than 1ms, and the time efficiency is low The requirement that the time-consuming algorithm of the application scenario be as short as possible

Images