Vehicle Surround Visualization via Single Camera and Synthetic Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle surround visualization systems using cameras have limited opening angles, failing to accurately depict the close-range environment, and require multiple cameras, increasing complexity and cost, while lacking a seamless transition between real-time video and synthetic models.
Innovation Solution
A method combining real-time video images with synthetically generated environment models, using a single video camera to determine obstacle distances and positions, allowing for a two-part display that integrates real-time video within the camera's range and synthetic models outside, enabling a comprehensive view with perspective smoothing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple cameras are used to capture the entire surrounding area, then the coverage area is improved, but the device complexity and manufacturing cost increase
Solution Approach 1:
The patent combines the functions of multiple cameras into a single camera by integrating real-time video capture with synthetic environment model generation. The single camera captures video images while distance sensors (ultrasonic, laser, or optical) provide depth information to construct a comprehensive 3D environment model that compensates for the camera's limited field of view, achieving full surrounding coverage without multiplying camera components.
Solution Approach 2:
The single video camera serves multiple functions: it captures real-time video images for direct display and provides visual reference data for generating the synthetic environment model. Distance sensors simultaneously perform both obstacle detection and provide spatial structure information. This multi-functionality allows one camera to effectively replace multiple cameras while maintaining comprehensive environmental awareness.
2Device complexity
If a single camera with limited opening angle is used, then the device complexity is reduced, but the close-range detection capability deteriorates
Solution Approach 1:
The system applies different quality characteristics to different spatial zones: the single camera provides high-quality visual information for distant objects within its field of view, while distance sensors provide precise measurement data for close-range obstacles. The synthetic environment model integrates these different quality sources, assigning appropriate detail levels to near and far regions, thereby achieving accurate close-range detection without requiring a camera array.
Solution Approach 2:
The synthetic environment model acts as an intermediary that bridges the gap between the single camera's limited visual capture and the need for comprehensive close-range detection. Distance sensor data serves as intermediate information, filling in the spatial details that the camera cannot capture directly, especially in the close-range zone where the camera's limited opening angle creates blind spots.
3Ease of operation
If only real-time video images are displayed, then the information intuitiveness is improved, but the coverage area is limited
Solution Approach 1:
The display is segmented into two complementary parts: real-time video images showing the actual visual appearance of distant objects provide intuitive information, while the synthetic environment model generated from distance sensor data fills in the close-range area that the camera cannot capture. This segmentation allows each component to excel at its strengths while collectively providing comprehensive coverage.
Solution Approach 2:
The system transitions from a two-dimensional camera image plane to a three-dimensional environment model that encompasses the entire surrounding space. The synthetic model adds the missing spatial dimension (close-range area outside camera view) while maintaining the intuitive visual representation, allowing the driver to perceive both distant visual details and near-field obstacles in a unified 3D spatial context.
4Area of stationary object
If synthetic environment models are used for full coverage, then the coverage area is improved, but the information realism deteriorates
Solution Approach 1:
The system applies synthetic model generation partially, only for the regions outside the camera's field of view. Within the camera's detection range, authentic real-time video images are displayed without synthetic overlay, preserving complete visual realism. The synthetic model is applied excessively only where necessary (outside camera range), providing coverage without compromising the realism of the captured visual information.
Data Source
Figure 1~3
Figure 4
Figure 5
AI summary
The invention relates to a method for visualizing the surroundings of a vehicle (1), comprising the following steps: determining and saving a current distance between the vehicle (1) and obstacles (6, 60, 61, 62, 63, 64, 65, 66) present in the surroundings of the vehicle by means of at least one sensor; determining and saving a current position of the vehicle; calculating an at least two-dimensional surroundings model from the saved data; calculating a virtual view (30) of the surroundings model from a selected virtual observer position (3); recording a video representation of at least one part of the surroundings by means of at least one video camera (2) and integrating the video representation (30) into the virtual view (30); and outputting the virtual view having the integrated video representation to a driver of the vehicle (1).