USV Proximity Recognition Using Multi-Image Collision Risk Estimation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Unmanned surface vehicles (USVs) face challenges in navigating through unpredictable sea conditions and reducing collision risks due to limited visibility and large inertial forces, which require advanced situational awareness systems.

Innovation Solution

A multi-image-based vessel proximity situation recognition system that uses multiple cameras and navigation sensors on USVs to detect obstacles and provide remote situational awareness of collision risks, utilizing image analysis, navigation data, and fuzzy inference for collision risk estimation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multiple cameras and sensors are used to improve situational awareness, then collision detection capability is improved, but device complexity increases

Engineering Contradiction:
Improvecollision detection capabilityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system divides the monitoring task into multiple specialized sensors (cameras, LIDAR, radar, AIS receivers) positioned at different locations on the USV. Each sensor handles specific detection tasks, and their data is processed separately before being integrated by the collision risk estimation unit, allowing improved detection capability while managing complexity through functional segmentation

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system merges data from multiple heterogeneous sensors (visual cameras, LIDAR point clouds, radar signals, AIS communications) into a unified collision risk assessment. The estimation unit integrates these diverse data streams to produce comprehensive situational awareness, achieving reliable collision detection by combining complementary sensor information

Inventive Principle:
Principle #5Merging (Combining)

2Reliability

If real-time monitoring and tracking are performed to improve collision avoidance, then safety is improved, but information processing requirements increase

Engineering Contradiction:
ImprovesafetyVSAvoidinformation processing load
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The system performs preliminary processing of sensor data to identify and track potential collision risks before they become immediate threats. The detection unit continuously monitors and tracks objects in advance, allowing the estimation unit to calculate collision risks proactively, which improves safety while distributing information processing load over time rather than requiring peak processing power at critical moments

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements continuous feedback loops where detection results feed into tracking, which feeds into collision risk estimation, which then informs updated detection priorities. This closed-loop information flow optimizes processing by focusing computational resources on high-risk targets identified through feedback from previous analysis cycles, improving safety while managing information processing requirements

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP4187347B1System and method for multi-image-based vessel proximity situation recognition support
Publication Date: 2025.01.29 KOREA INSTITUTE OF OCEAN SCIENCE & TECHNOLOGY
  • EP4187347B1 patent drawingFigure 1
  • EP4187347B1 patent drawingFigure 2
  • EP4187347B1 patent drawingFigure 3

AI summary

Proposed is a system and method for multi-image-based vessel proximity situation recognition support. The system for multi-image-based vessel proximity situation recognition support includes an unmanned surface vehicle (USV) configured to detect and track surrounding objects by monitoring surroundings using surrounding images and navigation sensors, and a remote navigation control device configured to support proximity situation recognition of the unmanned surface vehicle according to detection of the surrounding objects, wherein the unmanned surface vehicle may include an image acquisition unit, a navigation sensor unit, and a detection unit.