3D Scene Image Transmission Using Geometry-Only Feature Transfer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D scene reconstruction methods require large bandwidth resources due to the transmission of high-quality images from multiple cameras, leading to inefficiencies in data transmission and potential delays.
Innovation Solution
Disentangle geometry and appearance features from target images, transmitting only the geometry feature when quality requirements are met, and updating the appearance feature when necessary to reduce data transmission and improve efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high-quality images from multiple cameras are transmitted for 3D scene reconstruction, then the reconstruction quality is improved, but the bandwidth resource consumption increases
Solution Approach 1:
The patent segments the image data into two distinct feature types: geometry features (shape, structure, spatial relationships) and appearance features (colors, textures, materials). This segmentation allows selective transmission where only geometry features are sent to the terminal device, while appearance features are synthesized locally through neural rendering, thereby reducing bandwidth consumption while preserving reconstruction quality
Solution Approach 2:
The patent extracts and transmits only the essential geometry features from the original images, separating them from the appearance features. By taking out only the necessary geometric information and using neural radiance fields to synthesize appearance on-demand at the terminal, the system achieves high-quality 3D reconstruction with significantly reduced data transmission requirements
2Measurement precision
If more data is transmitted to improve 3D scene reconstruction quality, then the reconstruction accuracy is improved, but the transmission time increases
Solution Approach 1:
The patent extracts only the critical geometry features from the full image data and transmits them to the terminal device. This extraction of essential information maintains reconstruction accuracy while dramatically reducing transmission time, as geometry data is much smaller than complete high-quality images
Solution Approach 2:
The system performs preliminary processing on the server side to extract and prepare geometry features before transmission. By pre-processing the data to isolate only the necessary geometric information, the system minimizes transmission time while ensuring that all essential data for accurate 3D reconstruction is included
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
The present disclosure proposes a method and apparatus for image transmission for three-dimensional (3D) scene reconstruction. A target image may be obtained. A geometry feature and an appearance feature may be disentangled from the target image. An intermediate image may be reconstructed based on the geometry feature and a reference appearance feature. A difference between the intermediate image and the target image may be determined. The geometry feature may be transmitted to a receiving device for 3D scene reconstruction in response to determining that the difference is lower than a predetermined threshold.