A two-dimensional code image deblurring method based on adaptive composite prior

By employing an adaptive composite prior method, which combines multi-scale decomposition and significant edge region constraints, the deblurring problem of QR code images in dynamic scenes is solved, achieving efficient recovery of the module structure and edge information of the QR code and improving the decoding success rate.

CN122415381APending Publication Date: 2026-07-17CHENGDU YIGE MACHINERY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHENGDU YIGE MACHINERY CO LTD
Filing Date
2026-03-02
Publication Date
2026-07-17

Smart Images

  • Figure CN122415381A_ABST
    Figure CN122415381A_ABST
Patent Text Reader

Abstract

本发明涉及计算机视觉和数字图像处理技术领域，尤其涉及一种基于自适应复合先验的二维码图像去模糊方法；包括：首先，对模糊二维码图像进行多尺度金字塔分解；其次，在每一尺度上，基于像素聚类识别显著边缘区域，对其进行细分并施加极值约束，生成强化结构后的模糊图像；然后，构建自适应复合先验优化模型，采用交替优化算法迭代求解模型；迭代更新模糊核与潜在清晰图像，最终输出恢复的清晰二维码。本发明通过显著边缘区域的极值约束与自适应梯度先验的结合，能有效恢复二维码的锐利边缘和规整模块结构，显著减少振铃伪影，最终恢复的图像易于被标准解码器准确识别，具有较高的实用价值。
Need to check novelty before this filing date? Find Prior Art