LED Wall Correction for Virtual Production Backgrounds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques fail to produce a natural background image when using an LED wall with multiple panels, as they are designed for single-panel displays and do not account for the directional characteristics of LED elements, leading to unnatural background images during virtual production.
Innovation Solution
An image processing apparatus that acquires position and orientation information of the image capturing apparatus, calculates correction coefficients for each display element based on the display position, and corrects pixel values to ensure consistent luminance across the LED wall, thereby maintaining a natural background image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If image processing techniques designed for single-panel displays are used on LED walls with multiple panels, then the existing correction methods can be applied, but unnatural background images are displayed due to directional characteristics of LED elements and panel variations
Solution Approach 1:
The patent divides the LED wall into multiple individual LED panels and further into discrete display elements (LED elements). Each panel and display element is independently characterized and corrected. The system acquires luminance characteristics for each panel separately and applies individual correction coefficients to each display element, transforming the multi-panel system into manageable independent units that can be processed using existing single-panel correction techniques while achieving overall natural background images.
Solution Approach 2:
The patent applies local quality correction by calculating and applying unique correction coefficients to each display element based on its specific position, panel location, and directional characteristics. The correction amount varies locally across different panels and display elements to compensate for luminance variations, viewing angle dependencies, and panel-specific characteristics, ensuring each local region contributes to a natural overall background image.
2Area of stationary object
If LED walls with multiple panels are used to provide large display area, then the display area is increased, but directional characteristics of LED elements cause luminance to vary with viewing angle resulting in unnatural background images
Solution Approach 1:
The patent changes the luminance parameter of each display element dynamically based on the camera's position and orientation. By acquiring the camera's position information and calculating viewing angles, the system adjusts the luminance output of each LED element to compensate for directional characteristics. This parameter adjustment ensures consistent luminance perception across the entire multi-panel display area regardless of the camera's viewing angle.
Solution Approach 2:
The system uses feedback from camera position detection to continuously adjust display luminance. The image capturing apparatus's position and orientation are monitored, and this information feeds back to the image processing apparatus, which then recalculates and applies appropriate correction coefficients to maintain luminance consistency across the LED wall as the camera moves or changes angle.
3Manufacturing precision
If correction coefficients are calculated for each display element based on camera position and orientation, then luminance consistency is achieved, but processing complexity increases
Solution Approach 1:
The patent performs preliminary characterization of each LED panel and display element before actual virtual production. Luminance characteristics, viewing angle dependencies, and panel-specific parameters are pre-measured and stored. During operation, the system only needs to retrieve these pre-acquired characteristics and apply corrections based on camera position, significantly reducing real-time processing complexity while maintaining high luminance consistency.
Data Source
AI summary
An image processing apparatus comprises an acquisition unit configured to acquire information relating to position and orientation of an image capturing apparatus configured to capture an image of a subject with an image displayed on a display apparatus as a background, a calculation unit configured to calculate a correction coefficient corresponding to each display element included in the display apparatus, based on display position of the display apparatus and the information acquired by the acquisition unit, and a correction unit configured to correct, by the correction coefficient calculated by the calculation unit, a pixel value of each pixel in an image displayed on the display apparatus.