Autonomous Vehicle Perception with ROI Cropping for Long-Range Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles, particularly semi-trailer trucks, face limitations in perception range, often unable to detect objects at distances beyond 300 meters, which can lead to unsafe maneuvers or collisions due to insufficient image processing capabilities.
Innovation Solution
The implementation of a long-distance perception system that uses cameras to obtain images, processes them through a picture-in-picture module to crop regions of interest, and employs machine learning for object detection, along with depth estimation to determine the 3D position of objects up to 1000 meters away, utilizing terrain maps and camera pose information for accurate localization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image processing is performed on full-resolution images to detect objects at long distances, then object detection capability is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent divides the image processing task into two stages: first, a quick scan at lower resolution to identify potential regions of interest; second, focused processing only on those specific regions at full resolution. This segmentation approach maintains object detection capability while significantly reducing overall computational complexity by avoiding processing of entire high-resolution images.
Solution Approach 2:
The patent applies partial action by performing complete high-resolution processing only on portions of the image that contain objects of interest, rather than processing the entire image at full resolution. This allows the system to achieve sufficient detection precision for critical regions while reducing total computational load through selective processing.
2Reliability
If the perception range is extended to detect objects at 1000 meters, then safety is improved, but the system complexity increases due to additional processing requirements
Solution Approach 1:
The patent performs preliminary low-resolution scanning of the entire field of view to identify potential objects at long distances before committing resources to detailed analysis. This preliminary action enables the system to extend perception range to 1000 meters while managing complexity by filtering out non-relevant areas early in the processing pipeline.
Solution Approach 2:
The patent introduces an intermediary processing stage that bridges between low-resolution long-range detection and high-resolution detailed analysis. This intermediary stage identifies candidate regions and prepares them for further processing, enabling extended perception range without directly connecting the complexity of full-resolution processing across the entire field of view.
3Measurement precision
If high-resolution images are processed for accurate object localization, then positioning precision is improved, but processing speed decreases
Solution Approach 1:
The patent segments the processing workflow into rapid low-resolution screening followed by targeted high-resolution analysis only for regions containing objects of interest. This maintains positioning precision for detected objects while improving overall processing speed by avoiding high-resolution processing of empty or irrelevant image areas.
Solution Approach 2:
The patent applies partial action by using high-resolution processing only partially, specifically for localized regions where objects are detected, rather than processing the entire high-resolution image. This achieves sufficient positioning precision for critical objects while significantly reducing total processing time through selective application of computationally intensive operations.
Data Source
AI summary
Image processing techniques are described to obtain an image from a camera located on a vehicle while the vehicle is being driven, cropping a portion of the obtained image corresponding to a region of interest, detecting an object in the cropped portion, adding a bounding box around the detected object, determining position(s) of reference point(s) on the bounding box, and determining a location of the detected object in a spatial region where the vehicle is being driven based on the determined one or more positions of the second set of one or more reference points on the bounding box.