Novel View Image Control for Thin Object Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing novel view synthesis techniques face issues with inaccurate depth estimation and occlusion region filling, leading to side effects such as thin objects disappearing and limited viewpoint movement paths, resulting in distorted or incomplete stereoscopic images.
Innovation Solution
An electronic apparatus and method that identifies foreground and background regions using depth maps, predicts side effects based on object thickness and density, and controls viewpoint movement paths to minimize these effects by refining depth maps and adjusting occlusion regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If viewpoint movement path is expanded to provide more comprehensive stereoscopic images, then image completeness is improved, but thin objects may disappear and boundary distortion occurs
Solution Approach 1:
The system performs preliminary depth map refinement and thin object detection before viewpoint movement. By pre-identifying thin objects and refining depth boundaries in advance, the system prepares compensation data that will be applied during viewpoint transformation to prevent boundary distortion and object disappearance.
Solution Approach 2:
The system applies different processing strategies to different regions of the image. Thin object regions receive specialized depth refinement and viewpoint movement control, while other regions use standard novel view synthesis. This localized quality adjustment ensures boundary accuracy is maintained where needed without limiting overall viewpoint movement.
2Measurement precision
If depth map refinement is performed to improve object detection accuracy, then measurement precision is improved, but processing time increases
Solution Approach 1:
The system applies depth map refinement selectively only to regions containing thin objects or ambiguous boundaries, rather than processing the entire depth map uniformly. This localized refinement approach maintains measurement precision where critical while significantly reducing overall processing time.
Solution Approach 2:
The system performs rapid preliminary depth estimation first, then applies detailed refinement only to identified critical regions. This two-stage approach achieves high measurement precision for thin objects without the computational cost of refining the entire depth map.
3Reliability
If viewpoint movement path is restricted to prevent thin object disappearance, then object preservation is improved, but viewpoint flexibility deteriorates
Solution Approach 1:
The system dynamically adjusts viewpoint movement parameters based on detected thin object characteristics. For each thin object, the system calculates safe viewpoint ranges and adjusts movement paths accordingly, allowing maximum flexibility within preservation constraints rather than applying uniform restrictions.
Solution Approach 2:
The system uses detected thin object information as feedback to adaptively control viewpoint movement. By continuously monitoring object positions and depths, the system adjusts viewpoint paths in real-time to avoid configurations that would cause thin object disappearance while maintaining flexibility for comprehensive imaging.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
An electronic apparatus is disclosed. The electronic apparatus includes: a memory storing at least one instruction and at least one processor, comprising processing circuitry, individually and/or collectively, configured to: identify a foreground region and a background region included in an input image based on a depth map corresponding to the input image, and generate a novel view image by converting a viewpoint based on the foreground region; identify side effect prediction information including at least one from among whether an object of less than or equal to a specified thickness is included in the input image or an object density degree based on the depth map, and generate the novel view image by controlling a viewpoint movement path based on the identified side effect prediction information.