Virtual Driving Path Boundary Selection for Obstructed Lane Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating complex road environments where lane lines are obstructed or difficult to detect, leading to instability in generating a driving path.
Innovation Solution
An electronic device equipped with a sensing device and image obtainer processes data to generate multiple virtual driving paths, with a controller determining the actual driving path based on sensed data, image data, and position information, ensuring stability even in obstructed conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the vehicle uses sensor data and image data to generate a single virtual driving path, then the driving path generation is simple, but the stability of the driving path is reduced in complicated road environments
Solution Approach 1:
The patent divides the driving path generation into multiple virtual driving paths (first, second, and third virtual driving paths) with different boundaries. Each virtual driving path is generated based on different combinations of sensor data and image data, allowing the system to segment the decision-making process and select the most appropriate path for the current road environment.
Solution Approach 2:
The patent changes the boundary parameters of virtual driving paths by generating multiple paths with different boundary positions. The controller adjusts the boundary parameters based on the comparison between sensor data and image data, selecting the virtual driving path with the boundary closest to the vehicle to ensure stability in complicated road environments.
2Reliability
If the vehicle generates multiple virtual driving paths with different boundaries, then the stability of driving path is improved, but the processing complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-generating multiple virtual driving paths with different boundaries before making the final driving decision. This allows the system to prepare multiple options in advance and quickly select the most appropriate path when environmental conditions change, improving response time and stability.
Solution Approach 2:
The patent implements feedback by continuously comparing sensor data with image data and adjusting the selection of virtual driving paths based on this comparison. The controller uses the feedback from the environment recognition system to determine which virtual driving path boundary is closest to the vehicle, ensuring adaptive and stable path selection.
3Ease of operation
If the vehicle relies on lane line detection for path generation, then the path generation is straightforward, but the system fails when lane lines are blocked by foreign objects
Solution Approach 1:
The patent introduces an intermediary by using virtual driving paths as intermediate representations between raw sensor/image data and the final driving decision. These virtual paths act as mediators that can be adjusted and selected based on the reliability of lane line detection, allowing the system to bypass blocked lane lines by selecting alternative virtual paths with boundaries determined by other sensors or preset values.
Solution Approach 2:
The patent applies dynamics by making the virtual driving path boundaries adjustable and selectable rather than fixed. The system dynamically changes the boundary positions of virtual driving paths based on environmental conditions, allowing the vehicle to adapt to obstructed lane lines by selecting virtual paths with boundaries that are not blocked by foreign objects.
Data Source
AI summary
Provided is an electronic device including: a sensing device selected from a group including a radar and a lidar and installed in a vehicle to have a sensing zone directed to outside of the vehicle, the sensing device configured to obtain sensing data about an object; an image obtainer installed in the vehicle to have a field of view directed to the outside of the vehicle, the sensing device configured to obtain image data; and a controller including at least one processor configured to process the sensing data obtained by the sensing device and the image data obtained by the image obtainer, wherein the controller generates a first virtual driving path and a second virtual driving path based on the sensing data and the image data, and when a first boundary of the first virtual driving path and a second boundary of the second virtual driving path are located at different positions, provides a virtual driving path having a boundary closest to the vehicle between the first virtual driving path and the second virtual driving path as an actual driving path.


