Anti-interference method and device for intelligent multi-span greenhouse substrate weighing system
By integrating visual morphological features with weight time-series data, the system identifies and compensates for non-physiological interference, solving the data distortion problem caused by the tipping of vines in climbing crops and achieving high reliability and accuracy of the intelligent multi-span greenhouse substrate weighing system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-25
- Publication Date
- 2026-07-03
AI Technical Summary
Existing intelligent multi-span greenhouse substrate weighing systems suffer from data distortion due to non-physiological interference caused by the falling of vines in climbing crops. Traditional filtering algorithms and fixed threshold methods are unable to effectively distinguish between physiological and non-physiological interference, leading to inaccurate irrigation decisions.
By acquiring visual morphological features and weight time-series data of vine crops, fusing multimodal data representation, identifying and compensating for non-physiological interference, generating corrected weight time-series data, and using a visual transformer and a spatiotemporal dual-stream Transformer network for data fusion, and combining the water absorption characteristics of vine crops and a neural constant differential equation model for interference compensation.
It overcomes the blind spots of a single sensor, improves the reliability of weighing data, avoids false alarms of physiological signals, and ensures the accuracy of irrigation decisions.
Smart Images

Figure CN122329463A_ABST