3D Object Mesh Rendering on Mobile Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users often purchase items online without seeing them in person, leading to potential dissatisfaction due to size or style discrepancies, as photographs do not accurately represent the item's appearance in a physical environment.
Innovation Solution
A method for generating three-dimensional representations of objects using captured images from multiple views, creating an object mesh structure, and dynamically rendering synthetic images for arbitrary views by weighting and combining subset images, with caching and prefetching techniques to enhance rendering speed and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If photorealistic rendering of objects is provided to help customers visualize items in their environment, then purchasing decision accuracy is improved, but rendering time and computational resources increase
Solution Approach 1:
The system pre-renders images from multiple views of the object before the user actually needs them. When a user views an object in augmented reality, the controller has already generated images from various angles and positions, so the images are ready for immediate display without delay. This preliminary preparation resolves the contradiction by providing photorealistic accuracy without requiring time-consuming rendering during the actual user interaction.
Solution Approach 2:
The system dynamically adjusts the rendering process based on user interaction. As users move their devices to view objects from different angles, the controller detects these movements and selectively renders images only for the views that are currently needed. This dynamic approach ensures high visual accuracy for the specific viewpoint being observed while minimizing unnecessary rendering computations and time consumption for other angles.
2Measurement precision
If three-dimensional representations are generated using multiple captured views to enable arbitrary view rendering, then visualization accuracy is improved, but image processing complexity increases
Solution Approach 1:
The system divides the object into multiple discrete views or image slices captured from different angles. Each view is processed and stored separately as an independent image element. When rendering an arbitrary view, the controller selectively combines only the relevant segmented views needed for that specific angle, rather than processing the entire object model. This segmentation reduces processing complexity by breaking down the complex task of three-dimensional rendering into manageable, pre-processed image components.
3Adaptability or versatility
If synthetic images are created by weighting and combining subset images for arbitrary views, then rendering flexibility is improved, but computational requirements increase
Solution Approach 1:
The system applies partial weighting and combining operations only to the specific subset of images needed for the current arbitrary view, rather than processing all captured views. When a user requests a view from a particular angle, the controller identifies and combines only the relevant subset of pre-captured images that contribute to that view, applying weighting factors selectively. This partial action approach maintains rendering flexibility for any arbitrary angle while significantly reducing computational energy by avoiding unnecessary processing of irrelevant image subsets.
Data Source
AI summary
Images of various views of objects can be captured. An object mesh structure can be created based at least in part on the object images. The object mesh structure represents the three-dimensional shape of the object. Alpha masks indicating which pixels are associated with the object can be used to refine the object mesh structure. A request can be made to view the object from an arbitrary viewpoint which differs from the viewpoints associated with the captured images. A subset of the captured images can be used to create a synthetic image. Different weights can be assigned to the captured image to render a synthetic image that represents the view from the arbitrary viewpoint selected. The input images for the synthetic image can be prefetched, or loaded into memory before the arbitrary view is requested. The images can also be cached for future use or to avoid reloading them for another synthetic image.