Vehicle Image Distance Measurement Using Feature Point Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting front vehicles using vehicle shadows or machine learning fail when the lower portion of the vehicle is not included in the driving image, especially at close distances, posing a risk for accurate inter-vehicle distance measurement and autonomous driving.
Innovation Solution
The method employs feature point tracking to detect and measure inter-vehicle distance by identifying feature points in vehicle images, using machine learning or deep learning when possible, and optical flow to track these points when the lower portion is not visible, calculating distance based on feature point changes and vehicle size estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If shadow-based or learning-based detection methods are used, then detection performance is improved when vehicle lower portion is visible, but detection fails when the lower portion is not included in the driving image at close distances
Solution Approach 1:
The system dynamically switches between different detection methods based on the detection situation. When the lower portion of the vehicle is visible, shadow-based or learning-based methods are used; when it is not visible (close-distance scenario), feature point tracking is activated. This dynamic adaptation resolves the contradiction by making the detection system versatile across different driving conditions while maintaining high accuracy in each scenario.
Solution Approach 2:
Feature points serve as an intermediary element that bridges the gap between different detection methods. When conventional methods fail due to incomplete vehicle visibility, the system uses feature points extracted from the visible upper portion of the vehicle to track and measure distance. This intermediary approach enables continuous detection capability even when the lower vehicle portion is not visible.
2Measurement precision
If the distance between own vehicle and front vehicle becomes very close, then the lower portion of the front vehicle is not included in the driving image, but accurate distance measurement becomes more critical for collision prevention
Solution Approach 1:
Instead of relying on the lower portion of the vehicle for detection (which is not visible at close distances), the system inverts the approach by using feature points from the visible upper portion of the vehicle. This inversion allows the system to perform accurate distance measurement in close-distance scenarios where conventional detection methods fail.
Solution Approach 2:
The system transitions from two-dimensional vehicle region detection to tracking feature points across multiple frames in the time dimension. By extracting feature points from the visible vehicle portion and tracking their movement across frames, the system can measure distance accurately even when the vehicle occupies a small area in the image due to close proximity.
Data Source
AI summary
A method for measuring an inter-vehicle distance includes acquiring driving image photographed by a photographing device of a first vehicle which is being driven; detecting a second vehicle from the acquired driving image; detecting first feature points of a second vehicle region in a first frame corresponding to a frame in which the second frame is detected before a frame in which the second vehicle is not detected among a plurality of frames constituting the driving image, when the second vehicle is not detected from the driving image; detecting second feature points in a second frame corresponding to a current frame by tracking the detected first feature points; calculating a feature point change value between the first feature points and the second feature points; and calculating an inter-vehicle distance from the photographing device of the first vehicle to the second vehicle based on the calculated feature point change value.