Driving Skill Evaluation Using Curve Merging and Kernel Density Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing driving skill evaluation methods struggle to accurately assess a driver's skills due to variations in curve detection and evaluation processes, leading to inconsistent evaluation results.
Innovation Solution
A driving skill evaluation method and system that perform a detection process to identify curves based on vehicle traveling data, treating adjacent curves as one if the time or distance between them is less than a predetermined amount, and an evaluation process that assesses driving skills by comparing kernel density estimation images generated from the traveling data with reference images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple curves detected in close proximity are treated as separate curves, then the curve detection process is simple, but the evaluation accuracy deteriorates due to inconsistent curve section definitions
Solution Approach 1:
The patent applies the merging principle by combining multiple curves detected in close proximity into a single unified curve section. When the time or distance between consecutive curves falls below a predetermined threshold, the system merges these curves and treats them as one continuous curve section for evaluation purposes. This resolves the contradiction by improving evaluation accuracy through consistent curve section definition while managing detection complexity through automated merging logic.
Solution Approach 2:
The patent utilizes parameter changes by introducing a predetermined threshold parameter for time or distance between curves. This parameter serves as a criterion to determine whether to merge or separate detected curves. By changing the evaluation approach based on this parameter threshold, the system achieves both accurate curve section definition and manageable detection complexity through rule-based decision making.
2Productivity
If curve detection is performed without merging adjacent curves, then the detection process is faster, but the evaluation consistency deteriorates
Solution Approach 1:
The patent applies preliminary action by performing curve merging operations during the curve detection phase itself, rather than as a separate post-processing step. The system proactively identifies and merges adjacent curves based on the predetermined time or distance threshold before the evaluation process begins. This ensures evaluation consistency is established upfront while maintaining efficient processing through integrated detection and merging operations.
3Reliability
If subjective evaluation methods are used, then the evaluation process is simple, but the objectivity and reliability deteriorate
Solution Approach 1:
The patent replaces subjective human evaluation with an automated objective evaluation system that uses kernel density estimation and image processing. Instead of relying on human judgment, the system generates kernel density estimation images from traveling data and compares them against reference images using automated image processing algorithms. This substitution of mechanical/algorithmic evaluation for subjective human evaluation significantly improves objectivity and reliability while managing system complexity through established computational methods.
Solution Approach 2:
The patent creates visual representations (kernel density estimation images) that copy or represent the underlying traveling data in a comparable format. By transforming raw traveling data into visual image formats that can be systematically compared against reference images, the system enables objective evaluation while maintaining a manageable processing approach through visual data representation.
Data Source
AI summary
A driving skill evaluation method according to one embodiment of the disclosure includes: performing a detection process of detecting a curve based on traveling data of a vehicle; and performing an evaluation process of evaluating a driving skill of a driver of the vehicle, based on the traveling data at the curve. The detection process includes detecting a first curve and, when a time or a distance after an end of the first curve until detection of a second curve is less than a predetermined amount, treating the first curve and the second curve as one curve.