基于结构稳定概率测度表示的视觉识别方法、系统及设备

By constructing a visual recognition method based on structural stability probability measure representation, and combining semantic features and structural tensor matrix, the problem of insufficient stability of visual recognition in complex environments in existing technologies is solved, and higher recognition accuracy and robustness are achieved.

CN122049910BActive Publication Date: 2026-07-17HUNAN INST OF INFORMATION TECH
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HUNAN INST OF INFORMATION TECH
Filing Date
2026-04-17
Publication Date
2026-07-17

Smart Images

  • Figure CN122049910B_ABST
    Figure CN122049910B_ABST
Patent Text Reader

Abstract

本发明公开了一种基于结构稳定概率测度表示的视觉识别方法、系统及设备，方法包括：获取待识别图像并预处理；利用预训练深度神经网络提取语义特征向量；计算图像各局部区域的结构张量矩阵及结构相干性指标；对局部结构信息进行空间统计汇聚获得全局结构表征；基于所述语义特征向量确定概率分布均值向量，并根据所述全局结构表征生成调制后的协方差矩阵，以构建待识别图像的概率测度表示；通过计算与预构建类别原型概率测度表示之间的概率分布距离，确定识别结果。本发明将图像结构相干性信息融入概率分布的不确定性建模中，使特征距离度量更具结构感知性，从而有效提升了在遮挡、模糊等不确定场景下的识别鲁棒性与准确性。
Need to check novelty before this filing date? Find Prior Art