Multimodal perception integrated robot control system and method for fruit trees
By integrating RGB images, depth point clouds, six-dimensional force information, and tactile information into a multimodal perception and integrated operation robot system, combined with a tree digital twin module and a task planning module, the system solves the problems of low hardware reuse rate and disconnect between perception and decision-making in orchard robot systems for fruit thinning, girdling, and bud pinching operations. This achieves global optimization and consistency of orchard operations and improves the level of intelligence.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN WEIXIA ROBOT CO LTD
- Filing Date
- 2026-04-30
- Publication Date
- 2026-07-17
AI Technical Summary
Existing orchard robot systems suffer from problems such as low hardware reuse rate, disconnect between perception and decision-making, lack of force and tactile feedback, difficulty in coping with dynamic disturbances, and lack of quality review and parameter self-learning after operation in fruit thinning, girdling and bud pinching operations, resulting in insufficient consistency and intelligence level of operation.
The system employs a multimodal perception module that integrates RGB images, depth point clouds, six-dimensional force information, and tactile information. Combined with a tree digital twin module and a task planning module, it enables multimodal perception and integrated operation of fruit trees. The system uses a vision-force fusion control module for online correction, generates fusion control commands, and drives the robotic arm to complete precision operations.
It has achieved overall optimization and consistency in orchard operations, improved equipment utilization, enhanced the intelligence and precision of orchard management, and reduced reliance on manual labor and operational damage.
Smart Images
Figure CN122401397A_ABST