Fisheye Image Focus Peaking After Peripheral Distortion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image pickup apparatuses struggle with accurately determining focus in the peripheral parts of fisheye images, particularly in VR imaging, due to significant distortion that confuses the user's perception of focus state.
Innovation Solution
The apparatus performs predetermined transformation processing on captured images to correct distortion, applies focus peaking processing to transformed images, and displays a combined image with superimposed edge information to assist focusing, especially in peripheral areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If focus peaking processing is performed on the original fisheye image, then high-frequency components are extracted to highlight contours, but distortion in peripheral areas causes inaccurate focus determination
Solution Approach 1:
The image processing is divided into two distinct segments: first, transformation processing converts the fisheye image to a perspective projection image to correct distortion; second, focus peaking processing is applied to this transformed image. This segmentation allows each processing stage to optimize for its specific purpose without the interference of distortion artifacts.
Solution Approach 2:
The transformation processing is performed as a preliminary action before focus peaking processing. By pre-correcting the distortion in the fisheye image through perspective projection transformation, the subsequent focus peaking operation works on already-corrected image data, ensuring accurate focus determination without being misled by peripheral distortion.
2Ease of operation
If the fisheye image is displayed as-is, then the user sees the original captured image, but focus state in peripheral areas cannot be properly recognized due to distortion
Solution Approach 1:
The display mode parameter is changed based on the imaging mode. When VR imaging is detected, the system automatically switches to displaying the transformed perspective projection image instead of the original fisheye image. This parameter change ensures that the displayed image accurately represents focus state without distortion interference, making focus adjustment easier and more precise.
3Measurement precision
If peaking processing is applied to correct the focus indication, then edge information is highlighted, but the distorted peripheral objects are incorrectly identified as in-focus
Solution Approach 1:
Distortion correction through transformation processing is performed as a preliminary action before focus peaking processing. This ensures that when peaking processing highlights edge information, the underlying geometric relationships are already correct, preventing false focus indications in distorted peripheral areas.
Solution Approach 2:
The transformed perspective projection image serves as an intermediary between the original fisheye image and the final focus indication. This intermediary representation eliminates distortion artifacts that would otherwise cause false focus indications, allowing accurate focus assessment without the harmful effects of peripheral distortion.
Data Source
AI summary
An image pickup apparatus includes an imaging unit configured to acquire a captured image, and a processor configured to perform predetermined transformation processing for correcting image distortion for at least one partial area of the captured image to generate a transformed image, perform focus peaking processing for at least one of the captured image and the transformed image to generate a focus-peaking image, generate a combined image of the at least one of the captured image and the transformed image, and the focus-peaking image, and control a display unit so as to display the combined image. In a case where the processor is set to perform the focus peaking processing for the transformed image to display the combined image, the processor causes the display unit to display the combined image of the transformed image and the focus-peaking image. The partial area includes a peripheral part of the captured image.