Method for identifying character information on medical film based on deep learning

Abstract

The invention relates to a method for identifying character information on a medical film based on deep learning, and belongs to the technical field of medical image processing. The method comprises the following steps: S1, inputting a film image of character information to be recognized into a feature extraction network, and obtaining a character-containing multi-scale feature map; S2, inputting the feature map obtained in the step S1 into a multidirectional window extraction network to obtain a suggestion box; S3, processing the suggestion box in the step S2 by using an improved suggestion box rotation module to obtain a text detection result; and S4, inputting the text detection result in the step S3 into a coding and decoding character recognition module to obtain a corresponding character recognition result on the film. According to the invention, automatic identification of the character information on the medical film is realized, the corresponding Chinese and English character information on the medical film is directly output, manual correction and verification are not needed, the efficiency is high, and the accuracy is also guaranteed.

Description

Technical field [0001] The present invention belongs to the technical field of medical image processing, character information relates to a method for the medical film depth study based on automatic identification. Background technique [0002] Patients in the hospital radiology imaging techniques done checking, image interpretation doctors to make the diagnosis, patient care by, respectively, to print medical film and diagnostic reports. However, because the text font information on medical film is small, the contrast is not high, the information in the medical technicians sorting patient medical and diagnostic reports bagging film to check whether the two correspond unity, need to be very carefully inefficient, extended patient waiting time, very prone to human error. [0003] At present, character information for identifying a medical imaging methods rely mainly on manual tag acquired text feature, the need for large sample database text, including text positioning, pre-treatm...

AI Technical Summary

Problems solved by technology

However, because the text information on the medical film is small and the contrast is not high, medical technicians need to check whether the information on the two is consistent when sorting out the medical film and the diagnosis report for the patient. Increased patient waiting time and is prone to human error

[0003] The existing medical image text information recognition methods mainly rely on manual marking to obtain text features. It is necessary to establish a large sample database of text, including text positioning, preprocessing, recognition, correction, and verification. Narrowing the scope of recognition, correcting and verifying the text, the manual workload is still large, and there is also the probability of human error

Images