Method for identifying inland river container ship and related device
By constructing a port space set and a semi-supervised self-training model, and utilizing the berthing behavior characteristics of inland waterway container ships, the accuracy and robustness issues in inland waterway container ship identification were solved, achieving efficient automatic identification results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- YIHAILAN (BEIJING) DATA TECH CO LTD
- Filing Date
- 2025-12-10
- Publication Date
- 2026-07-21
AI Technical Summary
Existing technologies struggle to achieve robust and accurate automatic identification of inland waterway container ships, especially under limited known sample conditions. Traditional methods suffer from high annotation costs, insufficient feature utilization, and weak generalization and interpretability.
By acquiring data from the Automatic Identification System (AIS) of vessels and the geographical information of the wharf in the target waters, a port spatial set is constructed to identify vessel berthing events, calculate berthing frequency and comprehensive score, and use a semi-supervised self-training model for self-expansion learning to generate pseudo-labels, thereby enabling the determination of the type of vessel to be identified.
In complex inland waterway scenarios, relying solely on publicly available data and limited samples, inland container ships can be identified efficiently and accurately, providing a crucial data foundation for shipping research and management, and improving the accuracy and robustness of identification.
Smart Images

Figure CN121598169B_ABST
Abstract
Citation Information
Patent Citations
Ship type inference system and method
CN113849902A
Method and system for identifying end port of vessel
WO2023056656A1