Text region detection method based on embedded space pixel clustering

Abstract

The invention relates to a text region detection method based on embedded space pixel clustering, and solves the problem of inaccurate detection results of irregular text regions in the prior art. The method comprises the steps of obtaining a first original size text region prediction map, a first center line region prediction map, a first kernel region prediction map, a first height prediction map and a first angle prediction map of a shape space of a to-be-recognized text image, wherein the to-be-recognized text image comprises at least one first irregular text region; based on the first original-size text region prediction map, the first center line region prediction map and the first kernel region prediction map, obtaining a first pixel clustering prediction map of the first embedding space, optimizing the first center line region prediction map to obtain an optimized first center line region prediction map of the shape space, and combining the first height prediction map and the first angle prediction map to obtain a text region position in the to-be-recognized text image. Accurate detection of the irregular text area is realized, and the detection accuracy is improved.

Description

technical field [0001] The invention relates to the technical field of text detection, in particular to a text region detection method based on embedded spatial pixel clustering. Background technique [0002] In recent years, with the continuous development of text detection technology, the performance of scene text detection is also improving. The starting point of text area detection is to combine meaningful text instances as much as possible. According to different detection methods, it can be divided into component-based and segmentation-based And regression-based text area detection, because the result of text detection is a meaningful text instance, it is beneficial for the recognition end to extract richer text information, and it widely exists in Chinese and English locations and signs, and has great application prospects. [0003] However, due to the diversity of text in the scene, the detection of arbitrary-shaped text regions is also facing greater challenges. Th...

AI Technical Summary

Problems solved by technology

[0005] In view of the above analysis, the embodiment of the present invention aims to provide a text region detection method bas

Images