Data processing method and device for instrument detection, equipment and storage medium
By automatically generating training data from power distribution room images and introducing an attention mechanism and a high-resolution detection head module, the problems of data scarcity and difficulty in identifying small targets in power distribution room scenarios are solved, and high-precision digital instrument detection is achieved.
Patent Information
- Application Number
- CN202511900475.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-16
- Publication Date
- 2026-05-01
AI Technical Summary
In the context of power distribution rooms, traditional manual inspections are inefficient and prone to errors. Deep learning target detection models are difficult to adapt to the scarcity of data and the difficulty in identifying small targets, resulting in insufficient detection accuracy and failing to meet the accuracy requirements of industrial-grade inspections.
By automatically generating digital characters and overlaying them onto images of real power distribution room backgrounds, a large-scale training dataset is constructed. An attention mechanism and a high-resolution detection head module are introduced on the YOLOv8 framework to improve the model's feature extraction capabilities and small object detection accuracy.
It achieves high-precision and robust detection of small-sized digital instruments in complex power distribution room environments, improves detection accuracy and model training efficiency, and provides a low-cost industrial vision inspection solution.