Real-time detection system for food contaminants based on smart sensors
By using a multimodal sensor array, a dynamic baseline prediction model, and a signal separation network, the signal coupling problem caused by matrix changes during food processing was solved, enabling real-time detection and cross-scenario adaptation of food contaminants, and improving detection accuracy and efficiency.
Patent Information
- Application Number
- CN202610410074.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-31
- Publication Date
- 2026-05-29
AI Technical Summary
Existing food contaminant detection systems suffer from signal coupling due to matrix changes during food processing, making it difficult to effectively separate trace contaminant signals. Furthermore, they are difficult to adapt quickly across different scenarios, resulting in insufficient detection accuracy and high deployment costs.
By employing a multimodal sensor array combined with a time-series graph attention network and a variable decoupled autoencoder, a dynamic baseline prediction model and a signal separation network are constructed to achieve real-time decoupling of contaminant signals and matrix change signals during food processing. Furthermore, cross-scenario adaptive calibration is achieved through a meta-learning framework.
It enables real-time identification and accurate quantification of trace contaminant signals during food processing, reducing system deployment costs and timelines, and improving the timeliness of contamination incident detection and the accuracy of response.
Smart Images

Figure CN122109460A_ABST