Monocular Vehicle Camera Distance Estimation for Moving Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the distance to a moving object using a vehicle camera require multiple cameras to create a stereoscopic image, limiting their application to stationary objects and increasing costs and installation space.
Innovation Solution
A method and device using a single camera capture two images at different times, detecting a moving object and a stationary object with a specific geometric relationship, determining an image size ratio, and calculating distances based on the stationary object's known distances and vehicle speed to infer the moving object's distance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple cameras are used to create a stereoscopic image for distance determination, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent segments the distance measurement task into two parts: using a single camera to capture images at different times, and using temporal separation to simulate the spatial separation that would normally require multiple cameras. This allows monocular distance estimation without stereoscopic cameras.
Solution Approach 2:
The patent transitions from spatial dimension (using multiple cameras at different positions) to temporal dimension (using a single camera at different times). By capturing images at two different time points and using the vehicle's movement between these times, the system achieves depth information without requiring spatial separation of cameras.
2Measurement precision
If multiple cameras are used for distance determination, then measurement precision is improved, but installation space increases
Solution Approach 1:
The patent divides the measurement function across time rather than space, allowing a single camera to perform what would normally require multiple cameras. This eliminates the need for additional mounting spaces for multiple cameras while maintaining distance determination capability.
Solution Approach 2:
By moving the measurement capability from the spatial domain to the temporal domain, the system requires only one camera installation position. The vehicle's movement through space over time provides the additional viewpoint information that would otherwise require a second camera.
3Device complexity
If a single camera is used to capture images at different times, then device complexity is reduced, but the ability to determine distance to moving objects is limited
Solution Approach 1:
The patent uses feedback from the vehicle's known movement (odometry data) between the two time points to compensate for the motion of captured objects. By knowing how the camera position changed, the system can calculate what the object positions should be and use this information to determine distances even when objects are moving.
Solution Approach 2:
The patent changes the approach from using spatial parameters (multiple camera positions) to using temporal parameters (time between captures) combined with vehicle motion parameters. By incorporating vehicle speed and direction data, the system adapts the measurement calculation to account for object movement during the capture interval.
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for camera-based determination of the distance (b_1,b_2) of a moving object (60) in the surroundings (30) of a motor vehicle (50), wherein images (10, 11) of the surroundings (30) are detected at different times (t1, t2) by means of a camera (2), wherein the moving object (60) and at least one stationary object (61) are identified in the surroundings images (10, 11) captured and one stationary object (61) in the first surroundings image (10) is selected depending on the recorded geometric relationship relative to the moving object (60), an image size ratio (k) between the moving object (60) and the stationary object (61) is determined, and an image size (y'_2) of the stationary object (61) in the second surroundings image (11) is determined, with distances (a_1,a_2) of the stationary object (61) and an object size (y) of the stationary object (61) being determined on the basis of the image sizes (y'_1,y'_2) determined of the stationary object (61) and a vehicle speed (v_Fzg), and distances (b_1,b_2) of the moving object (60) are determined on the basis of the image size ratio (k) determined, the distances (a_1,a_2) determined, and the object size (y) of the stationary object (61) determined. The invention further relates to an associated device (1).