一种运单识别方法和计算机可读存储介质
By using a multimodal large model and a custom Prompt design, the problems of low efficiency and poor accuracy in waybill information entry are solved. Intelligent structured recognition and data support of waybill information are realized, which adapts to the transportation and logistics needs in the material management of the construction industry and improves the informatization and intelligence level of construction sites.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GLODON CO LTD
- Filing Date
- 2025-07-21
- Publication Date
- 2026-07-17
AI Technical Summary
In the management of material entry and exit in the construction industry, existing technologies rely on manual input of waybill information, which is inefficient and prone to errors. General OCR technology is difficult to implement effectively in complex engineering scenarios and cannot adapt to the problems of non-fixed format and large field variability of waybill documents.
By employing a multimodal large model combined with a custom Prompt design, the system acquires waybill images, performs image quality assessment and rotation correction, labels field names, and uses the multimodal large model to identify key field information. It supports both custom and built-in prompt words to achieve intelligent structured recognition of waybills.
It improves the efficiency and accuracy of waybill information entry, adapts to diverse form structures, provides accurate and timely data support, and promotes the informatization and intelligentization of construction sites.
Smart Images
Figure CN120877294B_ABST