Method for segmenting handwritten characters of financial bill small writing amount

Abstract

The invention discloses a method for segmenting handwritten characters of a financial bill small writing amount. The method is characterized by comprising the following steps: 01, searching an optimalsegmentation threshold for an input grayscale image by utilizing a K-means algorithm; 02, performing transverse and longitudinal projection segmentation on a character sample to be identified by using a histogram method to obtain a preliminary character segmentation sequence; and 03, using the character sequence segmentation sequence to calculate the central probability, namely the correlation coefficient, between the two characters through a CRAFT algorithm, if the correlation coefficient between the number amount characters is large and the correlation coefficient between the grid verticallines and the handwritten characters to be small, excluding grid lines, screening the optimal character segmentation sequence through the CRAFT algorithm, and outputting the result.

Description

technical field [0001] The invention relates to the field of sheet paper medium processing, in particular to a method for segmenting handwritten characters at the lowercase amount of financial notes. Background technique [0002] At present, there are various kinds of character processing software products of various artificial intelligence companies on the market, especially in character segmentation and recognition technology, which are quite mature. However, the application effect in specific recognition scenarios such as lowercase amount and uppercase amount on financial bills is not ideal. [0003] The types of characters are divided into printed numbers (0~9), handwritten numbers, printed uppercase and lowercase English letters (a~w), handwritten uppercase and lowercase English characters, printed special symbols and handwritten special symbols (such as ¥, $), etc. . Printed character recognition generally uses a relatively simple three-layer fully connected neural n...

AI Technical Summary

Problems solved by technology

However, the shape and size of handwritten characters are often different due to individual handwriting, and characters such as continuous strokes and sticking are very common, especially for handwritten numbers at lowercase amounts on bills. The standard also lies in the grid of currency numbers on the outside, and the handwriting becomes blurred after long-term storage, and the ink is dim, which will also bring certain difficulties

[0005] For the segmentation of such characters, traditional horizontal and vertical projection algorithms and deep learning neural network segmentation methods are difficult to achieve good results. The grid at the amount is easily misidentified as other characters (such as the number 1), and some handwritten bills The numbers will exceed the range of the grid, which will bring great challenges to the effective segmentation of characters

Images