Chinese complex scene text detection and recognition method

Abstract

The invention discloses a Chinese complex scene text detection and recognition method, relates to the field of computer vision, and realizes quick detection and recognition of Chinese texts in a complex scene. The method comprises the following steps of S0, obtaining Chinese complex scene sample data, and dividing the Chinese complex scene sample data into sample images of a training set and a test set according to a proportion of 8: 2; S1, performing image preprocessing operation on a training set sample to enable the training set sample to serve as input of a training model; S2, performing character region feature vector extraction on the training set sample through an improved darknet-19 network; S3, inputting the training sample into a preset YOLOv2 model for training to obtain a textdetection and recognition model; and S4, inputting the test sample into the trained model for testing to obtain a final identified detection box and a classification result of the character instances.According to the method, detection and recognition tasks are integrated into a unified network framework, and the method has high text detection and recognition performance and is suitable for text detection and recognition in a Chinese complex scene.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to a Chinese complex scene text detection and recognition method. Background technique [0002] Texts have always played an important role in people's lives. Rich and precise information contained in text is very important for vision-based applications, such as: image retrieval, object localization, human-computer interaction, robot navigation, and industrial automation, etc. Automatic text detection provides a way to acquire and utilize text information in images and videos, and thus becomes a hot research topic in the fields of computer vision and document analysis. [0003] Text detection in natural scenes is both an important and extremely challenging task. Since text detection in natural scenes usually recognizes text in the scene in an open scene, factors such as illumination, angle, and distortion cause great interference to text detection, which seriously affects the accurac...

AI Technical Summary

Problems solved by technology

The former such as: Fast-RCNN, Faster-RCNN and R-FCN, these methods can achieve high accuracy, but run slowly

Images