Autonomous Vehicle Image Fusion for Long-Range Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous vehicle navigation systems have limited perception range, which can be insufficient for semi-trailer trucks to detect objects in time, leading to potential collisions, as they rely on conventional image processing techniques that only allow for object detection within a limited range of up to 300 meters.
Innovation Solution
The implementation of a long-distance perception system that fuses real-time image feeds from multiple cameras to enhance object detection capabilities, allowing for accurate 3D positioning and motion modeling of objects up to 1000 meters away, using a processor-executable method that includes cropping images, identifying common objects, and fusing metadata to generate unified output results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple cameras and conventional image processing techniques are used, then object detection capability is provided, but the perception range is limited to up to 300 meters
Solution Approach 1:
The patent segments the image processing task by cropping multiple images from different cameras to focus only on regions containing detected objects. This segmentation approach reduces the amount of data that needs to be processed while maintaining detection accuracy, thereby extending the effective perception range without proportionally increasing processing complexity
Solution Approach 2:
The patent merges metadata from multiple cropped images that contain information about the same object. By combining detection results, bounding boxes, and object attributes from multiple camera views, the system achieves extended perception range (up to 1000 meters) while managing complexity through intelligent data fusion rather than processing all raw image data
2Reliability
If conventional image processing is used for autonomous vehicle navigation, then real-time operation is achieved, but the detection range is insufficient for semi-trailer trucks
Solution Approach 1:
The patent performs preliminary actions by detecting objects in full images first, then cropping only the relevant regions containing those objects. This preliminary detection and cropping approach ensures that no potential objects are missed (maintaining detection reliability) while significantly reducing the data volume for subsequent processing (improving processing efficiency)
Solution Approach 2:
The patent extracts only the essential information from multiple camera images by cropping out regions containing detected objects and fusing only the relevant metadata. This extraction approach maintains detection reliability by preserving all object information while improving processing efficiency by eliminating redundant background data
Data Source
AI summary
Devices, systems and methods for fusing scenes from real-time image feeds from on-vehicle cameras in autonomous vehicles to reduce redundancy of the information processed to enable real-time autonomous operation are described. One example of a method for improving perception in an autonomous vehicle includes receiving a plurality of cropped images, wherein each of the plurality of cropped images comprises one or more bounding boxes that correspond to one or more objects in a corresponding cropped image; identifying, based on the metadata in the plurality of cropped images, a first bounding box in a first cropped image and a second bounding box in a second cropped image, wherein the first and second bounding boxes correspond to a common object; and fusing the metadata corresponding to the common object from the first cropped image and the second cropped image to generate an output result for the common object.