Multi-Camera Depth-of-Field Selection for 3D Shape Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for generating virtual viewpoint images using multiple imaging devices struggle when capturing objects outside the depth of field, leading to inappropriate processing of shape data representing three-dimensional shapes.
Innovation Solution
An information processing apparatus determines whether a specific object exists within the depth of field of each imaging device and performs processing only on images where the object is in focus, utilizing both wide and narrow depth of field cameras to generate high-quality virtual viewpoint images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a plurality of imaging devices is arranged to capture the entire field with wide depth of field, then the coverage area is improved, but the image resolution and focus quality for specific portions deteriorate
Solution Approach 1:
The imaging system is segmented into multiple imaging devices with different depth of field characteristics. Some devices capture wide-area images with shallow depth of field for high-resolution portion capture, while others capture entire field images with deep depth of field. This segmentation allows the system to simultaneously achieve both wide coverage and high focus quality for specific portions.
Solution Approach 2:
The system performs partial action by selectively processing only those captured images where the specific object exists within the depth of field. Instead of using all captured images, the system identifies and utilizes only the subset of images that meet the focus quality criterion, thereby achieving high-resolution reconstruction without being constrained by images outside the depth of field.
2Quantity of substance
If captured images with objects outside depth of field are used for shape data generation, then the quantity of data is improved, but the accuracy of three-dimensional shape representation deteriorates
Solution Approach 1:
The system implements feedback by determining whether a specific object exists within the depth of field of each imaging device before using its captured image for shape data generation. This feedback mechanism ensures that only images with adequate focus quality are utilized, maintaining high three-dimensional shape accuracy while still leveraging the quantity of available captured images that meet the criterion.
3Productivity
If all captured images are processed for virtual viewpoint image generation, then the productivity is improved, but the image quality deteriorates due to inclusion of out-of-focus images
Solution Approach 1:
The system performs partial action by selectively processing only those captured images where the specific object exists within the depth of field. Instead of uniformly processing all captured images, it identifies and processes only the relevant subset, thereby maintaining high processing efficiency while ensuring that only high-quality in-focus images contribute to the virtual viewpoint image generation.
Data Source
AI summary
A plurality of images obtained by image capturing by a plurality of cameras and camera parameters are obtained, whether or not an object of interest exists within a depth of field of each camera is determined based on the camera parameters, and processing relating to shape data representing a three-dimensional shape of the object of interest is performed based on an image of the camera for which it has been determined that the object of interest exists within the depth of field.