Vehicle Self-Positioning Using Patterned Light and Feature Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-position calculating methods for vehicles face difficulties in accurately distinguishing feature points from patterned light beams projected on the road surface, leading to errors in calculating vehicle movement, especially when feature points are detected in the same area as the projected patterned light beam or far from it.
Innovation Solution
A self-position calculating apparatus and method that uses a light projector and camera to project and capture a grid-patterned light beam, employing a binarization process to extract the patterned light beam and detect feature points within a specifically set feature-point detection region adjacent to but distinct from the patterned light beam areas, allowing for accurate identification and calculation of vehicle position.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If feature points are detected in the same area as the projected patterned light beam, then the detection area is maximized, but it becomes difficult to distinguish between the patterned light beam and the feature points
Solution Approach 1:
The patent divides the image processing into two distinct segments: first extracting the patterned light beam area using binarization, then detecting feature points only in regions outside this extracted area. This segmentation allows maximization of the detection area while avoiding confusion between light beam and feature points by processing them in separate spatial zones.
Solution Approach 2:
The patent extracts the patterned light beam area from the captured image using binarization processing before feature point detection. By taking out and isolating the light beam region, the system can then search for feature points exclusively in the remaining areas, eliminating the distinguishing difficulty while maintaining large detection coverage.
2Difficulty of detecting and measuring
If feature points are detected in an area away from the projected patterned light beam, then the distinction between light beam and feature points is clear, but the error in calculating amounts of movements becomes larger
Solution Approach 1:
By segmenting the image into light beam areas and non-light beam areas, the patent enables feature point detection in regions that are clearly distinguished from the light beam while still maintaining proximity for accurate movement calculation. The segmentation creates optimal detection zones that balance distinction clarity with measurement precision.
3Measurement precision
If a binarization process is applied to extract the patterned light beam, then the light beam area is clearly identified, but the complexity of the processing increases
Solution Approach 1:
The binarization process serves as an intermediary step that simplifies the complex task of distinguishing the patterned light beam from the background and feature points. By converting the image to binary values, the system creates a clear mediator representation that makes subsequent feature point detection easier despite the added processing step.
Data Source
Figure 1
Figure 2
Figure 3(a)~3(b)
AI summary
A self-position calculating apparatus includes: a light projector 11 configured to project a patterned light beam onto a road surface around a vehicle; an image capturing unit 12 installed in the vehicle, and configured to capture and obtain an image of the road surface around the vehicle covering an area of the projected patterned light beam; an orientation angle calculator 22 configured to calculate an orientation angle of the vehicle relative to the road surface from a position of the patterned light beam on the image obtained by the image capturing unit 12; a feature point detector 23 configured to set a feature-point detection region surrounding the area of the projected patterned light beam on the image obtained by the image capturing unit 12, and to detect multiple feature points on the road surface within the feature-point detection region; an orientation change amount calculator 24 configured to calculate an amount of change in the orientation of the vehicle based on temporal changes in the multiple feature points on the road surface; and a self-position calculator 25 configured to calculate a current position and a current orientation angle of the vehicle by adding the amount of change in the orientation to an initial position and an initial orientation angle of the vehicle relative to the road surface.