Multi-line character paper form OCR (Optical Character Recognition) method based on semantic analysis

Abstract

The invention discloses a semantic analysis-based multi-line character paper table OCR (optical character recognition) method. According to the method, firstly, cell division is carried out on a to-be-digitalized paper table image, and then text content in the table is identified through a neural network. And based on semantic smoothness calculation, judging whether the recognition result is accurate or not. For out-of-order error recognition results, firstly, the row number and the column number of characters in original cells are extracted through a projection method, then a matrix is constructed, the error recognition results are sequentially filled into the matrix, then transposition operation is conducted on the matrix, the characters are output in order, and then a positive-order recognition result can be obtained. According to the method, automatic judgment and adjustment under the condition that multi-line character recognition in the paper form is out of order according to line / column recognition can be realized. The workload caused by manual proofreading is reduced, and the OCR intelligent degree is improved.

Description

technical field [0001] The invention belongs to the technical field of pattern recognition, and in particular relates to a multi-line text paper form OCR method based on semantic analysis. Background technique [0002] Paper forms store a lot of important information. With the continuous development of science and technology, the digitization of paper forms in certain fields has become an inevitable trend, such as cadre files, company employee files, etc. Digitizing these form data is beneficial to optimize management. The main method of digitizing paper forms is to use OCR for recognition. The full name of OCR is Optical Character Recognition. It refers to a process of inputting pictures and texts into computers and converting them into editable texts. The accuracy of OCR to recognize paper form information often depends on the complexity of the form frame. The simpler the form, the higher the accuracy of content recognition; the more complex the form, the lower the accura...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem in the prior art that the multi-line characters in the table are recognized out of order by row / column and cannot be automatically adjusted, the present invention proposes a multi-line text paper form OCR method based on semantic analysis, combined with the calculation of semantic fluency, through The matrix reorders the out-of-order text

Images