Sheet identifying device and method

Abstract

A paper identification device and method that can accurately determine the type or authenticity of paper by reducing the influence of paper creases and stains and the influence of watermark pattern relative to the positional deviation of the watermark area. If the transmissive photosensor (8) takes pictures of the paper, the watermark area extraction part (12) extracts the image data of the watermark area, and the watermark area center of gravity calculation part (13) calculates the center of gravity of the watermark area, and the watermark pattern shifts The amount calculation unit (14) calculates the position shift amount of the watermark pattern based on the center of gravity. The recognition target image extraction unit (15) extracts the recognition target image data based on the position shift amount, and the crease and stain detection unit (17) calculates the recognition target image data. The difference between the data and the true sample is detected to detect an image showing scratches and smudges, and the pattern matching unit (19) matches the image data of the recognition object whose influence of scratches and stains have been corrected and the real sample by pattern matching to identify the type or authenticity of the paper.

Description

Technical field [0001] The present invention relates to a paper recognition device and method, and more particularly, to a paper recognition device and method that extracts characteristics of paper from an image, and determines the type and authenticity of the paper. Background technique [0002] Generally, the identification of the type and authenticity of papers such as banknotes, stamps, or merchandise coupons is performed by using magnetic sensors or optical sensors to extract the characteristics of the papers inputted by the user through magnetic or light extraction. [0003] The optical characteristics of the paper are extracted by using the light sensor. By using the transmissive light sensor or the reflective light sensor, the pattern, size and direction of the paper are extracted, and their image patterns are obtained. The image patterns obtained by comparing with each type The standard graphics of genuine coupons can identify the type and authenticity of the input paper...

AI Technical Summary

Problems solved by technology

However, in the curve 342, due to the influence of creases and stains (331, 332, 333) etc. heavily, there is a possibility that the presence or absence of the watermark pattern 23 is judged incorrectly.

[0009] That is, in the conventional watermark pattern detection method, it is difficult to accurately determine the type or authenticity of the paper because it is easily affected by the state of the paper such as creases and stains.

[0010] In addition, in the conventional watermark pattern detection method, it is difficult to accurately determine the type of paper or whether it is true or false due to the influence of the positional offset of the watermark pattern relative to the watermark area.

Images