Aircraft Door Detection for Automated Boarding Bridge Docking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing passenger boarding bridge systems struggle to precisely detect the entrance of various types of aircraft due to varying aircraft sizes and gate arrangements, making it difficult to automate the docking process.
Innovation Solution
A detection system that includes a camera and an image processor to generate search area images by demarcating predetermined areas on captured images, shifting these areas along the aircraft axis, and performing projective transformations to enhance precision, using multiple cameras and adjusting image-capturing directions to ensure comprehensive coverage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single fixed search area is used for entrance detection, then the detection process is simple, but the system cannot adapt to various aircraft types and gate arrangements
Solution Approach 1:
The search area is dynamically adjusted and shifted along the aircraft axis direction based on the captured image content and aircraft type recognition. The system automatically determines the appropriate search area position and size for different aircraft configurations, transforming a static detection method into a dynamic adaptive one that handles various scenarios without manual reconfiguration
Solution Approach 2:
The detection process is divided into multiple stages: first identifying the aircraft type and gate arrangement, then segmenting the captured image into relevant search areas along the aircraft axis. This segmentation allows the system to focus computational resources on specific regions of interest while maintaining overall system simplicity
2Measurement precision
If the search area covers the entire captured image, then all possible entrance positions are checked, but the detection precision and efficiency decrease
Solution Approach 1:
Instead of uniformly processing the entire captured image, the system identifies and prioritizes specific local regions along the aircraft axis where entrances are most likely to be positioned. The search area is concentrated in these high-probability regions, improving detection precision while reducing processing time compared to full-image analysis
3Reliability
If multiple cameras are used to capture images from different directions, then the coverage and detection accuracy improve, but the system complexity and image processing load increase
Solution Approach 1:
The system uses a single camera that performs multiple functions by capturing images from different orientations. The camera rotates to capture the aircraft from various angles, and the same captured image is then processed to identify entrances along the aircraft axis. This multi-functional approach achieves reliable detection without requiring multiple fixed cameras
Solution Approach 2:
The camera performs periodic rotation and image capture operations to systematically collect information from different directions. This periodic imaging sequence allows the system to build comprehensive spatial understanding of the aircraft entrance locations through sequential observations rather than simultaneous multi-camera capture
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A detection system includes: a passenger boarding bridge connected to a terminal building; a camera on the passenger boarding bridge; and an image processor that detects an entrance of an aircraft from a captured image captured by the camera. The image processor includes: a search area image generator that generates a search area image by demarcating a predetermined area including part of the captured image; and a search performer that performs a search to determine whether or not the entrance is present within the search area image. The search area image generator shifts the predetermined area on the captured image from one side toward the other side of the aircraft in a direction of an aircraft axis of the aircraft to sequentially generate the search area image. The search performer repeatedly performs the search while changing the search area image by using multiple search area images that are sequentially generated.