Light Field Image Reconstruction with Region-Specific Focus Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image pickup apparatuses cannot independently control background blur and depth of field, leading to issues such as part of the object being blurred when trying to achieve a beautiful background blur, especially when photographing fast-moving objects or objects with depth.
Innovation Solution
An image processing apparatus and image pickup apparatus that divide an input image into regions and apply different shooting conditions, such as virtual imaging plane position, virtual depth of field, or virtual exposure, to each region, allowing for independent control of background blur and depth of field through a lens array and image processing methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If a lens with shallow depth of field is used to achieve beautiful background blur, then background blur quality is improved, but the entire object cannot be photographed clearly when the object has depth
Solution Approach 1:
The image is divided into multiple regions (foreground region, middle ground region, background region) and different focus positions are applied to each region independently. This allows the foreground to be in focus while the background remains blurred, resolving the contradiction between object sharpness and background blur quality.
Solution Approach 2:
Different focus positions and depth of field characteristics are applied to different regions of the image. The foreground region uses one focus position for sharpness, while the background region uses another focus position for blur effect, allowing each region to have optimal quality for its intended purpose.
2Ease of operation
If multiple images are shot with changing focus position to perform blur control, then depth of field control is improved, but positions of moving objects differ between images making blur control difficult
Solution Approach 1:
All images with different focus positions are captured simultaneously in a single shot using a light field camera, rather than capturing multiple images sequentially. This preliminary capture of all focus information at once eliminates the problem of moving object position changes between shots, while still allowing post-capture depth of field control.
Solution Approach 2:
The patent transitions from sequential multi-shot photography to simultaneous light field capture, adding the dimension of angular information. This allows all focus positions to be captured at the same moment, preserving moving object positions while enabling post-processing depth of field control.
Data Source
AI summary
An image processing apparatus generates an output image using an input image including object space information, the image processing apparatus includes a region dividing unit configured to divide the input image into a first region and a second region, and a reconstructed image generating unit configured to reconstruct the input image to generate the output image, and the reconstructed image generating unit applies different shooting conditions from each other to the first region and the second region respectively.