Character segmentation and recognition method based on CTC deep neural network

Abstract

The invention provides a character segmentation and recognition method based on a CTC deep neural network. The method comprises the following steps: a1, extracting features of an input image by usinga CNN; a2, carrying out the CELL segmentation on the features extracted in a1, fixing the height and width of CELL, and determining the number of CELL by the length of the image; a3, directly segmenting and classifying each CELL of the determined features, and outputting segmentation signals; a4, calculating the loss between the real segmentation signal and the segmentation signal output by the model by using CTCLOSS, feeding back the loss condition and training the whole model; a5, segmenting the text by using the segmentation signal output in the step a3, carrying out CNN + softmax classification identification on a single character, mapping a real segmentation signal from the annotated text, and automatically solving the text alignment problem by using the CTCLOSS. According to the invention, the OCR recognition speed is improved, and the recognition optimization is targeted after the characters are cut into single characters, so the final precision is improved; a recognition framework is improved, and a recognition process is separated into character segmentation and single character recognition, so optimization can be separately carried out in a targeted manner.

Description

technical field [0001] The present invention relates to the technical field of text segmentation and recognition, in particular to a method for text segmentation and recognition based on a CTC deep neural network. Background technique [0002] OCR (Optical Character Recognition) is an image processing technology that detects, recognizes, and structures text in images. The current OCR technology is divided into three modules: detection, recognition, and structuring; there are two frameworks for detection and recognition, which are: 1. Single-character detection + single-character recognition framework, specifically expressed as the core task of the detection module is detection Each independent character area of ​​the image; the recognition module is responsible for text recognition for each cropped character area image. The basic framework of the existing recognition model is: CNN+softmax; 2. Text line detection + entire line recognition framework, specifically It is stated...

AI Technical Summary

Problems solved by technology

[0004] The text information in the image content is structured based on templates and rule logic. The existing frameworks have certain disadvantages. Under the first framework above, the detection needs to mark the position of each character, and the cost of labeling is extremely high. At the same time, the difficulty of structuring is greatly increased. Therefore, the specific goal of general detection tasks is to detect text lines, rather than to detect individual characters independently; under the second framework, the recognition module takes a lot of time

Images