Map Information Processing Device for Occlusion Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aerial images may contain occlusions such as multistoried intersections, high-rise buildings, or bridges, which can hinder the extraction of road information, leading to incomplete or erroneous road information generation for autonomous driving and driver assistance systems.
Innovation Solution
A map information processing device that extracts road areas and demarcation lines from aerial images, estimates occlusion positions, determines road continuity near occlusions, and generates accurate traffic lane information to ensure reliable high-precision map verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If road information is extracted from aerial images without considering occlusions, then the processing is simple and fast, but the reliability of road information decreases due to incomplete data in occluded areas
Solution Approach 1:
The occlusion estimation unit performs preliminary estimation of occlusion areas before road information extraction. By predicting where occlusions occur based on image characteristics and map information, the system prepares for potential data gaps in advance, allowing subsequent processing to focus only on critical areas and improving overall reliability without proportionally increasing complexity
Solution Approach 2:
The patent introduces map information as an intermediary resource to verify road continuity in occluded areas. Instead of directly inferring continuity from incomplete aerial image data, the system uses pre-existing map information as a mediator to confirm whether roads should continue through occluded regions, thereby improving reliability through cross-validation
2Area of stationary object
If the range of high-precision map is enlarged, then more roads are covered, but verification cost increases
Solution Approach 1:
The system performs self-verification by automatically comparing extracted road information against map information stored in the map information storage unit. This automated self-check mechanism eliminates the need for manual verification processes, allowing the map range to be expanded without proportionally increasing verification costs
Solution Approach 2:
The verification unit provides feedback by comparing generated road information with existing map data and identifying discrepancies. This feedback mechanism enables automated detection of errors and inconsistencies, reducing the need for costly manual verification while expanding map coverage
3Productivity
If demarcation lines are connected assuming continuity without occlusion verification, then traffic lane information is completed, but erroneous information may be included
Solution Approach 1:
Map information serves as an intermediary verification layer between raw aerial image extraction and final traffic lane information generation. The system uses this intermediate map data to validate whether assumed continuities are correct, preventing erroneous connections while maintaining processing efficiency
Solution Approach 2:
The patent replaces manual verification mechanisms with automated image processing and comparison algorithms. The occlusion estimation unit and verification unit automatically analyze image characteristics and compare data, substituting human judgment with computational processes that maintain accuracy while improving productivity
Data Source
AI summary
Provided is a map information processing device which contributes to the improvement of high-precision map verification performance by generating road information at a traffic lane level for a verification purpose by functioning as a means for extracting a road area corresponding to a road and a demarcation line corresponding to a traffic lane from an aerial image (for example, a ground photo photographed from an aircraft or a satellite), inferring the position of an object/structure (occlusion) which may cover the road, determining whether or not the demarcation line in the vicinity of the occlusion forms a continuous line by using the inferred occlusion position and demarcation line information, and next determining whether or not the traffic lane in the vicinity of the occlusion has continuity such that a misunderstanding about the continuity of the road when the presence of no occlusion is possible can be avoided to suppress the inclusion of incorrect information. This device generates traffic lane information which can be also used to generate appropriate traffic lane information for an area for which road information cannot be extracted due to an occlusion by using a satellite image/aerial photo for a map information verification purpose.