Ghost Elimination Using Brightness-Adjusted Underexposed Reference Frame
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In high dynamic scenes, moving objects often result in ghosting artifacts due to exposure frame intervals, leading to image blurring and loss of information during ghost elimination, as existing methods either add noise or discard highlight details.
Innovation Solution
A method involving acquiring three images with different exposures, increasing the brightness of the underexposed image, and using it as a reference frame for ghost elimination, along with histogram matching and format conversion, to align motion regions and retain image details.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If ghost elimination is performed using normal exposure image as reference frame, then processing speed is maintained, but ghosting artifacts occur in long-burst regions due to motion deviation
Solution Approach 1:
The patent applies preliminary action by adjusting the brightness of the underexposed image before using it as a reference frame for ghost elimination. This preprocessing step ensures that the reference frame has sufficient brightness information to accurately align motion regions, preventing ghosting artifacts from occurring during the ghost elimination process.
Solution Approach 2:
The patent changes the brightness parameter of the underexposed image to create a suitable reference frame. By adjusting the brightness parameter, the underexposed image becomes capable of serving as an effective reference for motion alignment, thereby improving ghost elimination accuracy without introducing ghosting artifacts.
2Device complexity
If underexposed image is used directly as reference frame, then processing complexity is reduced, but image details are lost due to insufficient brightness
Solution Approach 1:
The patent applies preliminary action by adjusting the brightness of the underexposed image before using it as a reference frame for ghost elimination. This preprocessing step ensures that the reference frame has sufficient brightness information to accurately align motion regions, preventing ghosting artifacts from occurring during the ghost elimination process.
Solution Approach 2:
The patent changes the brightness parameter of the underexposed image to create a suitable reference frame. By adjusting the brightness parameter, the underexposed image becomes capable of serving as an effective reference for motion alignment, thereby improving ghost elimination accuracy without introducing ghosting artifacts.
3Measurement precision
If brightness of underexposed image is increased, then motion region alignment is improved, but noise may be introduced during adjustment
Solution Approach 1:
The patent applies local quality by selectively adjusting the brightness of the underexposed image only in the regions where motion detection and alignment are needed. This localized brightness adjustment improves motion region alignment accuracy while minimizing the introduction of noise to other parts of the image.
Solution Approach 2:
The patent changes the brightness parameter of the underexposed image to create a suitable reference frame. By adjusting the brightness parameter, the underexposed image becomes capable of serving as an effective reference for motion alignment, thereby improving ghost elimination accuracy without introducing ghosting artifacts.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a method and device for processing an image, and a storage medium. A first image of a target object, a second image of the target object, and a third image of the target object are acquired (101). Exposure of the first image is less than exposure of the second image. The exposure of the second image is less than exposure of the third image. A fourth image is acquired by increasing brightness of the first image (102). A second reference frame is acquired by performing ghost elimination is performed (103) on the second image and taking the fourth image as a first reference frame. A fifth image is acquired (104) by performing ghost elimination on the third image based on the second reference frame.