Driving Skill Evaluation Using Kernel Density Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing driving skill evaluation methods lack accuracy in assessing a driver's skills, particularly in complex driving conditions, necessitating an improvement in evaluation precision.
Innovation Solution
A driving skill evaluation system that generates kernel density estimation images based on time-series data of vehicle parameters such as longitudinal acceleration rate and yaw angular velocity, comparing these images with reference images to determine driving proficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional evaluation methods using only longitudinal and lateral acceleration rates are used, then the evaluation process is simple, but the evaluation accuracy is insufficient
Solution Approach 1:
The patent transforms time-series driving parameter data into a two-dimensional kernel density estimation image, adding a visual dimension to the evaluation process. This allows complex multi-parameter data (longitudinal acceleration, lateral acceleration, steering angle, etc.) to be represented and compared in a graphical format, improving evaluation accuracy while maintaining manageable complexity through image-based analysis
2Measurement precision
If multiple parameters are used to improve evaluation comprehensiveness, then the evaluation becomes more accurate, but the data processing complexity increases
Solution Approach 1:
The patent merges multiple driving parameters (longitudinal acceleration rate, lateral acceleration rate, steering angle, etc.) into a single kernel density estimation image. By combining these diverse parameters into one unified visual representation, the system achieves comprehensive evaluation without proportionally increasing processing complexity, as the image integration consolidates what would otherwise require separate analysis of multiple data streams
Data Source
AI summary
A driving skill evaluation method according to one embodiment of the disclosure includes: generating a kernel density estimation image, based on time-series data of a first parameter corresponding to a direction change in a traveling direction of a vehicle, and time-series data of a second parameter indicating a square of a jerk in the traveling direction of the vehicle; and evaluating a driving skill of a driver of the vehicle, by comparing the kernel density estimation image with a reference image.