Text Detection Method and System Based on Cross-Level Feature Enhancement and Auxiliary Feature Guidance
By using cross-level feature enhancement and auxiliary feature guidance, the receptive field of low-level features and edge information enhanced by high-level features are dynamically adjusted, which solves the problem of detecting text with large scale differences and extreme aspect ratios in natural scene images, and achieves text detection with higher accuracy and efficiency.
CN120126110BActive Publication Date: 2026-05-26TIANJIN UNIV
2 Cites 0 Cited by
Patent Information
- Application Number
- CN202510175820.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-18
- Publication Date
- 2026-05-26
- Estimated Expiration
- 2045-02-18
AI Technical Summary
Technical Problem
Existing text detection methods struggle to effectively detect text with large scale differences and extreme aspect ratios in natural scene images, resulting in low detection accuracy and efficiency.
Method used
We employ a method of cross-level feature enhancement and auxiliary feature guidance. By dynamically adjusting the receptive field of low-level features and the differential convolution of high-level features, we enhance high-frequency edge information. Combined with attention guidance, we generate more accurate text bounding boxes.
Benefits of technology
It improves the detection accuracy and efficiency of text at different scales, avoids the problem of the same text being misdetected as multiple regions, and enhances the overall performance of the model.
✦ Generated by Eureka AI based on patent content.
Abstract
This invention discloses a text detection method and system based on cross-level feature enhancement and auxiliary feature guidance, belonging to the field of computer vision technology. The method includes the following steps: extracting feature maps of different levels from the image to be detected; for low-level features, dynamically adjusting the receptive field by adjusting the convolutional kernel size to obtain multi-scale information, generating enhanced features for low-level features; for high-level features, using differential convolution to enhance high-frequency information at text edges, generating enhanced features for high-level features; fusing features from different levels using cross-level feature enhancement to generate fused features; extracting semantic information from high-level features to generate auxiliary features; for the generated auxiliary features and fused features, using attention guidance to determine the attribution of text kernel pixels, obtaining attention-guided features; and performing text-to-non-text prediction on pixels to generate text bounding boxes. The scheme in this invention enables the network to more accurately segment text regions.
Need to check novelty before this filing date? Find Prior Art