Intelligent real-time monitoring method and system for precision seeding
By acquiring crop adaptation parameters, image processing, and improved YOLOv8s model detection in precision seeding, the problems of hill-sowing cycle attribution and seed status association were solved, enabling real-time monitoring and status determination of precision seeding, and improving counting accuracy and seeding quality evaluation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- QINGDAO UNIV OF TECH
- Filing Date
- 2026-05-20
- Publication Date
- 2026-07-17
AI Technical Summary
Existing precision seeding monitoring methods struggle to assign seeding cycles within continuous image frames, leading to inaccurate seeding counts. Furthermore, the lack of correlation between seed integrity and seeding quantity makes it difficult to synchronously reflect the seeding status.
By acquiring crop adaptation parameters, image acquisition is performed and hole-sowing cycle numbers are assigned. Combining image cropping, brightness correction, and sharpness judgment, an improved YOLOv8s target detection model is used to detect seed targets and identify integrity. Through cross-frame tracking and deduplication counting, hole-level status judgment is output.
It achieves accurate hill-sowing cycle assignment and deduplication counting for continuous image frames, which can reduce the repeated counting of the same physical seed, and outputs real-time monitoring results of precision sowing including hill-sowing cycle number, pass rate, missed sowing rate, re-sowing rate and damage rate.
Smart Images
Figure CN122397422A_ABST