Autonomous Driving Camera Blockage Detection for Small Occlusions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving systems face challenges in detecting camera field-of-view blockages, particularly when only a small portion is blocked, leading to inaccurate determinations and high false-positive rates due to limited information in single image frames.
Innovation Solution
An autonomous driving system that evaluates pixel values in captured images to identify candidate blocked regions, eliminates non-blockage causes, and monitors chronological changes to determine likely camera blockages, using spatial-temporal classifiers and local statistical features to generate an alarm signal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If single-image blur detection mechanism is used, then detection speed is improved, but false-positive rate increases
Solution Approach 1:
The patent divides the image into multiple processing units (e.g., grid segments) and evaluates each unit independently for blur characteristics. This segmentation allows the system to identify localized blockages more accurately without being misled by overall image blur, thereby reducing false positives while maintaining detection speed through parallel processing of segments.
Solution Approach 2:
The patent transitions from analyzing a single image frame to analyzing multiple consecutive image frames temporally. By examining the temporal consistency of blur patterns across frames, the system can distinguish between actual camera blockages (which persist across frames) and transient blur causes (which appear and disappear), significantly reducing false-positive rates while maintaining efficient detection.
2Device complexity
If single-image blur detection mechanism is used, then system complexity is reduced, but blockage detection accuracy for small portions deteriorates
Solution Approach 1:
By dividing the image into multiple processing units, the system can detect small blockages in specific segments without requiring complex analysis of the entire image. Each segment is evaluated independently, allowing the system to maintain low overall complexity while achieving high detection accuracy for localized blockages.
Solution Approach 2:
The patent applies blur detection selectively to specific processing units rather than uniformly across the entire image. This partial action approach allows the system to focus computational resources on regions where blockages are likely to occur, improving detection accuracy for small portions without significantly increasing overall system complexity.
3Measurement precision
If spatial-temporal analysis with multiple images is used, then blockage detection accuracy is improved, but processing time increases
Solution Approach 1:
The patent processes images by dividing them into segments and evaluating each segment independently. This segmentation enables parallel processing of multiple segments across multiple frames, reducing the overall processing time while maintaining the accuracy benefits of spatial-temporal analysis.
Solution Approach 2:
The system performs blur evaluation at periodic intervals across multiple consecutive frames rather than continuously analyzing every frame. This periodic action reduces processing time while still capturing sufficient temporal information to accurately distinguish persistent blockages from transient blur, maintaining high detection accuracy with reduced computational overhead.
Data Source
AI summary
Systems and methods for detecting blockages in images are described. An example method may include receiving a plurality of images captured by a camera installed on an apparatus. The method may include identifying one or more candidate blocked regions in the plurality of images. Each of the candidate blocked regions may contain image data caused by blockages in the camera's field-of-view. The method may further include assigning scores to the one or more candidate blocked regions based on relationships among the one or more candidate blocked regions in the plurality of images. In response to a determination that one of the scores is above a predetermined blockage threshold, the method may include generating an alarm signal for the apparatus.