Sensor Resolution Control for Adaptive Autonomous Driving Perception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to efficiently suppress the processing load of high-resolution data from external-environment sensors in vehicles, particularly when considering road types and structures, leading to increased arithmetic requirements.
Innovation Solution
A resolution control device that processes output data from external-environment sensors, dynamically controlling the resolution based on traveling status, such as road type and structure, by extracting an attention region and reducing the resolution of other regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution data from external-environment sensors is processed, then measurement precision is improved, but computing power requirement increases
Solution Approach 1:
The image processing is divided into two distinct paths: a first processing path for a first region of the image and a second processing path for a second region of the image. This segmentation allows different processing loads to be applied to different regions, reducing overall computing power requirements while maintaining necessary precision in critical areas.
Solution Approach 2:
Different processing qualities are applied to different regions of the image based on their importance. The first region (containing important information) receives high-resolution processing, while the second region receives lower-resolution processing. This local differentiation maintains measurement precision where needed while reducing computing power requirements in less critical areas.
2Productivity
If resolution of entire image is reduced, then processing load is suppressed, but important region recognition accuracy deteriorates
Solution Approach 1:
The system applies different processing qualities to different regions: the first region (containing important information for autonomous driving) is processed with high resolution to maintain recognition accuracy, while the second region is processed with lower resolution to reduce overall processing load. This selective approach ensures that productivity is improved without sacrificing accuracy in critical areas.
Solution Approach 2:
The image is segmented into multiple regions with different processing requirements. By separating the image into a first region requiring high processing and a second region tolerating lower processing, the system can suppress overall processing load while preserving recognition accuracy for important regions.
3Device complexity
If fixed processing method is used, then device complexity is reduced, but adaptability to different road situations deteriorates
Solution Approach 1:
The processing method dynamically adapts to different driving situations by determining the vehicle's traveling status (straight, left turn, right turn) and adjusting which regions receive high-resolution processing accordingly. This dynamic adaptation improves versatility without significantly increasing device complexity, as the system only needs to add status determination logic to the existing processing framework.
Data Source
AI summary
Provided is a resolution control device that suppresses a processing load of output data by dynamically controlling a resolution at time of processing the output data of an external-environment sensor based on a traveling status. The resolution control device processes output data of an external-environment sensor that recognizes a surrounding environment of a vehicle and outputs the processed output data to an automatic driving device. The resolution control device includes a traveling status determination unit that determines a traveling status of the vehicle, and a control unit that extracts an attention region of the output data based on the traveling status, reduces a resolution of a region other than the attention region, and then outputs the region to the automatic driving device.