2D Virtual Object Coverage Edges Without 3D Scene Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing methods require costly and computationally complex devices like lidar to reconstruct three-dimensional spaces for adding effects such as snow buildup and AR scenes, which is time-consuming, especially in high-complexity scenarios.
Innovation Solution
A method and apparatus that determine a coverage area and edge of a virtual object based on an image and a time variable, allowing the addition of a coverage effect without constructing a three-dimensional space, performed on a two-dimensional plane.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a lidar device is used to reconstruct three-dimensional space for adding coverage effects, then the real-space effect is improved, but the device cost and computational complexity increase
Solution Approach 1:
The patent extracts only the essential spatial information needed for coverage effects (coverage area and edge) from the complex three-dimensional reconstruction process. Instead of using lidar to build complete 3D models, the method extracts 2D coverage boundaries directly from images, eliminating the need for complex 3D spatial reconstruction while maintaining the visual effect of coverage.
Solution Approach 2:
The patent creates a simplified two-dimensional copy of the coverage effect instead of working with the full three-dimensional reality. By generating a 2D coverage area mask and edge from the image plane, the method replicates the visual appearance of coverage effects without requiring actual 3D spatial data, thus reducing device complexity while preserving the visual realism.
2Reliability
If a lidar device is used to reconstruct three-dimensional space for adding effects, then the coverage effect quality is improved, but the processing time increases
Solution Approach 1:
The method extracts only the critical coverage boundary information from images rather than performing complete three-dimensional reconstruction. By taking out just the coverage area and edge data needed for the visual effect, the processing time is significantly reduced while maintaining the quality of the coverage effect.
Solution Approach 2:
The patent transitions from three-dimensional spatial reconstruction to two-dimensional image plane processing. By working directly with 2D coverage masks and edges on the image plane instead of reconstructing full 3D spaces, the method achieves comparable visual效果的 quality with much faster processing speed.
3Measurement precision
If three-dimensional space reconstruction is performed to add virtual object effects, then the spatial accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent extracts the essential spatial boundaries (coverage area and edge) directly from two-dimensional images without performing complete three-dimensional reconstruction. This extraction approach maintains sufficient spatial accuracy for coverage effects while avoiding the high computational complexity of full 3D spatial modeling.
Solution Approach 2:
The method uses lightweight two-dimensional coverage masks and edge representations instead of heavy three-dimensional spatial models. These 2D structures are computationally inexpensive to process and discard after use, providing the necessary spatial accuracy for coverage effects without the burden of complex 3D data structures.
Data Source
Figure 1
Figure 2a~2b
Figure 2c~3a
AI summary
The present disclosure discloses a method and apparatus for processing an image, a device, and a medium. After the image to be processed is obtained, a coverage area corresponding to a virtual object in the image to be processed is determined, and a coverage edge of the virtual object is determined based on the coverage area and a time variable. Since a degree of coverage of the virtual object in the coverage area and the corresponding coverage edge change over time, the coverage edge of the virtual object needs to be determined based on the coverage area and the time variable. Further, a coverage effect of the virtual object is obtained based on the coverage area and the coverage edge. That is, after the coverage area and the coverage edge are determined, the virtual object is added to the corresponding coverage area. As such, with the technical solutions provided in the present disclosure, the addition of the coverage effect of the virtual object in the image to be processed has no need for a lidar to construct a three-dimensional space of a scene in the image to be processed, and the entire processing process is performed on a two-dimensional plane, thereby reducing computation and improving convenience.