To address the problem of underutilization of features in traditional micro-image recognition algorithms for Chinese herbal medicines due to the neglect of text data, a multimodal adaptive fusion method for micro-image recognition of Chinese herbal medicines was proposed. The problem of low recognition accuracy due to the large intra-class differences and small inter-class differences in the microscopic feature images of Chinese herbal medicines was addressed by collaboratively analyzing the image and text data. Firstly, a semantic-driven attention-based dynamic mask generator was introduced, which generated dynamic masks from text input and served as an effective constraint to mitigate misclassification between similar microscopic images in the final classification. Secondly, a microscopic image feature encoder was designed to extract both overall information and fine-grained features of microscopic images. Lastly, a self-attention mechanism was incorporated into the multimodal fusion module, where a text-driven network was utilized to filter image features, thereby reducing redundant information and enhancing classification accuracy. Experimental results show that recognition accuracy of 97.5% is achieved by the model, showing improvements of 7.7, 6.1, 4.3, and 2.9 percentage points compared to the Biformer-B, InceptionNeXt-B, EfficientFormer V2-L, and ConvNeXt V2-L models, respectively. In summary, the proposed multimodal adaptive fusion method effectively compensates for the feature defects of single-image input and provides a reliable technical solution for the high-precision recognition of microscopic images of traditional Chinese medicinal materials.
TANG Zhi-qiang, WANG Yi-ding.
Microscopic Image Recognition of Traditional Chinese Medicines Based on Multimodal Adaptive Fusion[J].
Science Technology and Engineering, 2026, 26(8): 3377-3385 DOI:10.12404/j.issn.1671-1815.2501894
ZhaoZ, LiY, ZhouL, et al. Prevention and treatment of COVID-19 using traditional chinese medicine: a review[J]. Phytomedicine, 2021, 85: 153308.
[2]
陈凯先, 张卫东. 中药现代化与中药创新[J]. 中国食品药品监管, 2022(8): 4-13.
[3]
ChengKaixian, ZhangWeidong. . Modernization and innovation of traditional Chinese medicine[J]. China Food & Drug Administration Magazine, 2022(8): 4-13.
YangJun. Research progress on the application of microscopic identification technology in the identification of Chinese medicinal materials[J]. Guangming Traditional Chinese Medicine, 2021, 36(16): 2828-2830.
TanChaoqun, WenChuanbiao, WuCunjie. Research on the identification of Chinese medicine decoction pieces based on image processing technology[J]. Lishizhen Medicine and Materia Medica Research, 2018, 29(7): 1706-1709.
ChengShiyan, LuWenbiao, WangFengmei. Image recognition of Chinese herbal pieces based on color matching template[J]. Chinese Journal of Experimental Traditional Medical Formulae, 2020, 26(6): 158-162.
WangYiding, YaoYi, LiYaoli, et al. Research on microscopic image recognition of Chinese medicinal materials power based on improved dynamicReLU and attention mechanism model[J]. Application Research of Computers, 2021, 38(9): 2861-2865, 2870.
WangYiding, SunXia, LiYaoli, et al. Study on detection method for microscopic images of powdered Chinese medicinal materials[J]. Computer Applications and Software, 2022, 39(4): 80-87.
[20]
CaoB, SunY, ZhuP, et al. Multi-modal gated mixture of local-to-global experts for dynamic image fusion[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. New York: IEEE, 2023: 23555-23564.
[21]
TolstikhinI O, HoulsbyN, KolesnikovA, et al. MLP-mixer: an all-MLP architecture for vision[J]. Advances in Neural Information Processing Systems, 2021, 34: 24261-24272.
[22]
WangW, XieE, LiX, et al. Pyramid vision Transformer: a versatile backbone for dense prediction without convolutions[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. New York: IEEE, 2021: 568-578.
[23]
DosovitskiyA, BeyerL, KolesnikovA, et al. An image is worth 16×16 words: transformers for image recognition at scale[J]. arXiv Preprint, arXiv: 2010.11929, 2020.
[24]
ZhuH, CaoG, ZhaoM, et al. Effective image tampering localization with multi-scale ConvNeXt feature fusion[J]. Journal of Visual Communication and Image Representation, 2024, 98: 103981.
[25]
RadfordA, KimJ W, HallacyC, et al. Learning transferable visual models from natural language supervision[C]// International Conference on Machine Learning. Dover: PMLR, 2021: 8748-8763.
[26]
GuoR, WeiJ, SunL, et al. A survey on image-text multimodal models[J]. arXiv Preprint, arXiv: 2309.15857, 2023.
[27]
XuJ, SunX, ZhangZ, et al. Understanding and improving layer normalization[J]. Advances in Neural Information Processing Systems, 2019, 32: 8-14.
[28]
KumarT, BrennanR, MileoA, et al. Image data augmentation approaches: a comprehensive survey and future directions[J]. IEEE Access, 2024, 12: 187536-187571.
WooS, DebnathS, HuR, et al. ConvNeXt v2: co-designing and scaling ConvNets with masked autoencoders[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2023: 16133-16142.
[31]
YuW, ZhouP, YanS, et al. InceptionNeXt: when inception meets ConvNeXt[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2024: 5672-5683.
[32]
WangA, ChenH, LinZ, et al. RepViT: revisiting mobile CNN from ViT perspective[C]// 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA: IEEE, 2024: 15909-15920.
[33]
ZhuL, WangX, KeZ, et al. BiFormer: vision transformer with bi-level routing attention[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2023: 10323-10333.
[34]
LiY, YuanG, WenY, et al. EfficientFormer: vision transformers at Mobilenet speed[J]. Advances in Neural Information Processing Systems, 2022, 35: 12934-12949.