Ship Safety Supervision With Deep-Learning Video Early Warning

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current surveillance video data processing on ships relies heavily on manual personnel judgment, leading to high operating costs, low efficiency, and security risks, with limitations in close-range and multi-angle supervision, real-time intelligent analysis, and lack of early warning.

Innovation Solution

An intelligent safety supervision system for ships incorporating a ship-side and shore-side supervision system, utilizing image acquisition, automatic recognition via deep learning, ship and shore servers for feature recognition, alarm modules, and communication modules to enable real-time online supervision and early warning, with secondary feature recognition and data processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual personnel judgment is used for surveillance video data processing, then the system can operate with existing infrastructure, but operating cost increases and efficiency decreases

Engineering Contradiction:

Improvesupervision efficiencyVSAvoidmanpower required

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The patent replaces manual personnel judgment with an intelligent recognition system that uses deep learning algorithms (convolutional neural networks) to automatically analyze surveillance video data. The system substitutes human cognitive processing with automated computational models that can process video streams in real-time, thereby increasing supervision efficiency while reducing the quantity of human manpower required.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service through automatic feature recognition and alarm generation. The intelligent recognition system independently processes surveillance data, identifies anomalies, and triggers alarms without requiring continuous human intervention. The ship-side server automatically performs feature recognition on video data and generates alarm indications that are transmitted to the shore-side system, allowing the system to monitor and alert itself autonomously.

Inventive Principle:

Principle #25Self-service

2Area of stationary object

If remote supervision is implemented, then coverage rate improves, but accuracy at close range and multi-angle supervision deteriorates

Engineering Contradiction:

Improvesupervision coverageVSAvoidclose-range supervision accuracy

Core Design Contradiction:

Area of stationary objectVSMeasurement precision

Solution Approach 1:

The patent divides the supervision system into two segments: a ship-side supervision system that handles local close-range monitoring with high-precision cameras and intelligent recognition, and a shore-side supervision system that provides remote centralized monitoring. The ship-side system processes video data locally to maintain accuracy for close-range supervision, while the shore-side system provides broader coverage through remote access to processed data and alarm information.

Inventive Principle:

Principle #1Segmentation

3Speed

If manual video checking is performed, then system complexity remains low, but real-time analysis capability and early warning deteriorate

Engineering Contradiction:

Improvereal-time analysis speedVSAvoidsystem complexity

Core Design Contradiction:

SpeedVSDevice complexity

Solution Approach 1:

The system performs preliminary action by pre-processing and feature recognition on the ship-side before data transmission to the shore-side system. The ship-side server automatically extracts features and generates alarm indications in advance, enabling real-time analysis capability. This preliminary processing allows the system to maintain speed for real-time detection while managing complexity through hierarchical data processing.

Inventive Principle:

Principle #10Preliminary action

4Productivity

If intelligent recognition system is deployed, then supervision efficiency improves, but data transmission stability in weak network environments deteriorates

Engineering Contradiction:

Improvesupervision efficiencyVSAvoiddata transmission stability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent extracts and processes critical data locally on the ship-side before transmission. The ship-side intelligent recognition system performs feature extraction and alarm generation locally, transmitting only essential alarm indication information to the shore-side system. This extraction of critical processing tasks to the local environment reduces the data transmission burden and improves reliability in weak network conditions while maintaining high supervision efficiency.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS12573186B2Intelligent safety supervision system applied to ship

Publication Date: 2026.03.10 CHENGDU UNIVERSITY OF TECHNOLOGY

AI summary

An intelligent safety supervision system applied to a ship is provided. An image acquisition module acquires high-definition images in real time. An automatic recognition module obtains ship dynamic and static data. A ship server performs feature recognition on the ship dynamic and static data to obtain a data processing result, transmits the ship dynamic and static data and the data processing result, and receives alarm indication information. An alarm module outputs an alarm. A ship client displays the data processing result, and determines whether to transmit the alarm indication information according to the data processing result. A communication module receives and transmits the ship dynamic and static data and the data processing result. A shore-side supervision system includes a ship safety supervision big data analysis platform for performing secondary feature recognition on the ship dynamic and static data, so as to obtain a secondary data processing result.