A tobacco foreign matter identification and rejection method and system based on multi-modal perception

By combining multimodal perception technology with multidimensional information fusion and confidence assessment, the problems of insufficient recognition capability and poor adaptability of traditional tobacco foreign object identification systems have been solved, achieving high-precision and low-error intelligent foreign object removal.

CN122397952APending Publication Date: 2026-07-17HEBEI BAISHA TOBACCO
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HEBEI BAISHA TOBACCO
Filing Date
2026-03-16
Publication Date
2026-07-17

Smart Images

  • Figure CN122397952A_ABST
    Figure CN122397952A_ABST
Patent Text Reader

Abstract

本发明公开了一种基于多模态感知的烟叶异物识别与剔除方法及系统，涉及工业自动化检测领域，包括：获取烟叶输送带上烟叶流的多模态图像数据，并进行预处理，得多模态图像块；进行多层次特征提取，并通过特征融合网络生成融合特征图；利用异物检测与分类模型进行异物区域的定位、分割以及类别预测，并输出各类别的预测置信度；根据异物区域的类别、预测置信度，进行异物风险等级评估与剔除决策，生成包含质量等级标识的决策指令；生成进行异物剔除的动作控制信号，并根据质量等级标识决定是否触发数据重采集或模型优化流程。其通过多模态数据融合、智能分级决策与优化流程，实现了对烟叶流中高准确率、低误剔率的高效异物剔除。
Need to check novelty before this filing date? Find Prior Art