Two-dimensional recursive network-based recognition method of Chinese text in natural scene images

Abstract

The invention discloses a two-dimensional recursive network-based recognition method of Chinese text in natural scene images. Firstly, a training sample set is acquired, and a neural network formed bysequentially connecting a deep convolutional network, a two-dimensional recursive network used for encoding, a two-dimensional recursive network used for decoding and a CTC model is trained; test samples are input into the trained deep convolutional network, and feature maps of the test samples are acquired; the feature maps of the test samples are input into the trained two-dimensional recursivenetwork, which is used for encoding, to obtain encoding feature maps of the test samples; the encoding feature maps of the test samples are input into the trained two-dimensional recursive network, which is used for decoding, to obtain a probability result of each commonly used Chinese character in each image of the test samples; and clustering searching processing is carried out, and finally, the overall Chinese text in the test samples is recognized. According to the method of the invention, space / time information and context information of the text images are fully utilized, the text imagepre-segmentation problem can be avoided, and recognition accuracy is improved.

Description

technical field [0001] The invention belongs to the field of image text analysis and recognition, in particular to a method for recognizing Chinese text in natural scene images based on a two-dimensional recursive network. Background technique [0002] Most of the human information is obtained through the visual system. The scene image obtained through the visual system not only contains rich visual information such as color, pattern, shape, position, texture, but also rich text information. The description of information by text has the characteristics of accuracy and validity, and text has very useful value in various computer vision applications. For example, in terms of image search, recognizing the text in the image will help us better classify and match the image; in terms of unmanned driving, recognizing the text information of traffic signs and other signs from natural scenes can assist driving. In today's rapid development of artificial intelligence, text recogniti...

AI Technical Summary

Problems solved by technology

However, this kind of method still has certain defects. For example, the distortion of the text in the image, such as rotation and transmission, requires a large number of sample training to enhance the recognition ability of the network. When recognizing a one-dimensional recursive network, it is necessary to first convert the two-dimensional feature map into a one-dimensional image. dimensional feature sequence

Images