Vehicle Lane Recognition Using Virtual Camera Object Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle control systems struggle to accurately recognize the position of objects in images captured by multiple cameras, leading to positional deviations of demarcation lines and objects, which can compromise traffic safety.
Innovation Solution
A vehicle control device that includes an imager, a recognizer, a driving controller, and a controller. The controller determines the belonging lane of an object by analyzing the positional relationship between the left and right demarcation lines and the edges of objects on a two-dimensional image captured by the imager, allowing for more accurate driving control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If coordinate transformation into bird's-eye view is performed for each camera image, then the field of view is expanded and multiple objects can be recognized, but positional deviation of demarcation lines and objects occurs reducing measurement precision
Solution Approach 1:
The patent divides the recognition process into two separate systems: one for detecting demarcation lines (lane markings) and another for detecting objects (vehicles, pedestrians). Each system processes images independently without coordinate transformation, maintaining measurement precision while still achieving comprehensive situational awareness through separate detection channels.
Solution Approach 2:
The patent introduces a virtual camera positioned at the host vehicle's location as an intermediary. This virtual camera captures a two-dimensional image that serves as a reference frame, allowing the system to determine belonging lanes and relative positions without performing coordinate transformations on actual camera images, thus preserving position accuracy.
2Adaptability or versatility
If coordinate transformation is applied to integrate multiple camera views, then comprehensive surroundings recognition is achieved, but positional relationship accuracy between demarcation lines and objects deteriorates
Solution Approach 1:
The patent segments the recognition task into independent demarcation line detection and object detection processes. By keeping these processes separate and avoiding coordinate transformation, the system maintains accurate positional relationships while still achieving comprehensive surroundings recognition through the integration of detection results from multiple cameras.
Solution Approach 2:
The patent creates a two-dimensional image from a virtual camera as a copy or representation of the actual three-dimensional scene. This two-dimensional image serves as a reference for determining belonging lanes and positional relationships, allowing the system to achieve comprehensive recognition without distorting actual object positions through coordinate transformation.
3Area of stationary object
If multiple cameras are used to capture images, then the coverage area increases, but the complexity of determining accurate lane belonging increases
Solution Approach 1:
The patent uses a virtual camera as an intermediary to simplify the lane determination process. The virtual camera's two-dimensional image provides a straightforward reference frame for determining which lane an object belongs to, eliminating the need for complex coordinate transformations and making the system more manageable despite using multiple cameras.
Data Source
AI summary
A vehicle control device according to an embodiment includes an imager configured to image surroundings of a host vehicle, a recognizer configured to recognize a surroundings situation of the host vehicle, a driving controller configured to control one or both of speed and steering of the host vehicle on the basis of a result of the recognition of the recognizer, and a controller configured to control the driving controller on the basis of imaging content of the imager, and the controller determines a belonging lane of the object on the basis of a positional relationship between left and right demarcation lines of a traveling lane of the host vehicle and edges of an object present around the host vehicle on a two-dimensional image captured by the imager.