一种基于深度学习的印刷品检测方法及系统
By combining multi-dimensional topological mapping matching and dynamic memory bank updates with multi-channel information entropy paths and non-equilibrium dissipation structures, the robustness and sensitivity issues of existing printed matter detection methods under complex textures and multi-scale information are solved, enabling accurate identification and real-time feedback of printed matter defects.
CN120912498BActive Publication Date: 2026-07-17ANHUI ZHEYINAN PREPRINT TECH CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ANHUI ZHEYINAN PREPRINT TECH CO LTD
- Filing Date
- 2025-06-08
- Publication Date
- 2026-07-17
Smart Images
Figure CN120912498B_ABST
Abstract
本发明属于印刷品检测技术领域，本发明公开了一种基于深度学习的印刷品检测方法及系统，包括采集印刷品图像数据并进行多维拓扑映射匹配，利用基于特征分布形态与多尺度曲率半径对比方法，计算印刷品图像特征匹配度；将匹配度结果存储于检测记忆库，通过基于重放记忆的记忆更新机制，结合检测记忆库，对检测记忆进行动态增强与修正；将匹配度结果按照香农信息理论进行多通道信息熵路径建模，通过特征分维压缩与熵增重整方法进行自适应特征传输优化，采用动态余量冗余重传机制进行重传验证，得到高质量特征数据；提升了印刷品缺陷检测的准确率和实时响应能力，适应复杂多变的印刷品检测需求。
Need to check novelty before this filing date? Find Prior Art