Neural Network Obstacle Detection for Autonomous Maritime Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current obstacle detection systems in autonomous ships are limited by inaccuracy and reliance on visual inspection, particularly due to GPS inaccuracy, AIS update periods, and radar limitations such as non-observation regions and noise, making it difficult to achieve reliable obstacle sensing for autonomous navigation.
Innovation Solution
A neural network-based method for image segmentation is employed to analyze marine images, identifying obstacles and their distances, which updates the neural network using error calculations and labelling data to improve obstacle detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If GPS, AIS, and radar systems are used for obstacle detection, then automated navigation is enabled, but detection accuracy is insufficient due to GPS inaccuracy, AIS update periods, radar non-observation regions, and noise
Solution Approach 1:
The patent combines multiple sensor systems (GPS, AIS, radar) with visual inspection capabilities into an integrated obstacle detection system. The neural network fuses data from these different sources to compensate for individual system limitations, achieving both automated navigation and improved detection accuracy by leveraging the complementary strengths of each sensor type.
Solution Approach 2:
The patent introduces a neural network as an intermediary processing layer between raw sensor data and navigation decisions. This neural network processes and fuses information from GPS, AIS, radar, and visual inputs, transforming multiple imperfect measurements into reliable obstacle detection results that enable autonomous navigation while overcoming individual sensor deficiencies.
2Measurement precision
If visual inspection is used to accurately detect obstacles, then detection accuracy improves, but operational complexity increases and autonomous operation becomes difficult
Solution Approach 1:
The patent implements a neural network-based system that performs automatic visual inspection and obstacle detection without human intervention. The system processes images, identifies obstacles, and integrates this information with navigation data autonomously, maintaining high detection accuracy while reducing operational complexity by eliminating the need for manual visual checking.
Solution Approach 2:
The patent replaces manual visual inspection with an automated neural network vision system. The neural network processes visual data and detects obstacles automatically, substituting the mechanical process of human visual checking with an electronic imaging and processing system that achieves comparable or superior accuracy with less operational complexity.
3Reliability
If multiple sensor systems are integrated for comprehensive obstacle detection, then detection reliability improves, but system complexity and computational requirements increase
Solution Approach 1:
The patent designs a neural network that serves multiple functions simultaneously: it processes data from GPS, AIS, radar, and visual sensors; performs obstacle detection; calculates distances; and provides output for navigation decisions. This multi-functional approach integrates multiple sensor systems while managing complexity through a unified processing architecture that handles diverse inputs through a single versatile system.
Data Source
AI summary
A method for situation awareness is provided. The method comprises: preparing a neural network trained by a learning set, wherein the learning set includes a plurality of maritime images and maritime information including object type information which includes a first type index for a vessel, a second type index for a water surface and a third type index for a ground surface, and distance level information which includes a first level index indicating that a distance is undefined, a second level index indicating a first distance range and a third level index indicating a second distance range greater than the first distance range; obtaining a target maritime image generated from a camera; and determining a distance of a target vessel based on the distance level index of the maritime information being outputted from the neural network which receives the target maritime image and having the first type index.