Related Experiment Video
Updated: May 18, 2026
05:48
Memorization-Based Training and Testing Paradigm for Robust Vocal Identity Recognition in Expressive Speech Using Event-Related Potentials Analysis
Published on: August 9, 2024
PHDReader: police handwritten document recognition method based on VLM with EI-LFT using FP-EESR and MFE-GLS
Dongxing Wang1,2, Zhiyang Huang3
1Department of Computer and Information Security Management, Fu Jian Police College, Fuzhou, 350007, China. wangdongxing85@163.com.
Scientific Reports
|May 16, 2026
Summary
This study introduces PHDReader, a novel method for police handwritten document recognition. It significantly improves accuracy and generalization compared to existing OCR and VLM models.
Area of Science:
- Computer Science
- Artificial Intelligence
- Pattern Recognition
Background:
- Police handwritten document recognition is crucial for public security.
- Existing methods face challenges in accuracy and adaptability.
Purpose of the Study:
- To develop an advanced method for police handwritten document recognition.
- To enhance the recognition performance, stability, and generalization ability of handwritten Chinese character recognition.
Main Methods:
- Utilized Fusion Preprocessing with Edge Enhancement for Super-Resolution Reconstruction (FP-EESR) for feature enhancement.
- Employed Multi-scale Feature Extraction using Global-Local-Semantic (MFE-GLS) for feature fusion.
- Applied Embedding Instruction & LoRA Fine-Tuning (EI-LFT) to adapt Qwen3VL for police document scenarios.
- Developed a specialized dataset for police handwritten document recognition and extraction.
Main Results:
- The proposed PHDReader method demonstrated superior recognition performance.
- PHDReader showed enhanced stability and generalization capabilities.
- Comparative experiments confirmed the effectiveness against advanced OCR and VLM models.
Conclusions:
- PHDReader offers a significant advancement in police handwritten document recognition.
- The method effectively addresses the challenges in recognizing complex handwritten police documents.
- This research contributes to improving the combat capability of public security organs through advanced AI.