Image Processing for Four-Corner Trailing Blur Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current imaging systems suffer from inconsistent image quality due to four-corner blur, which affects user experience and cannot be effectively addressed by hardware improvements or conventional image deblurring techniques, leading to hardware cost and image distortion issues.
Innovation Solution
A method that predicts trailing blur state information based on object position in an input image using a Coma prediction module, employing a two-level attention mechanism and dynamic convolution operations to adjust processing according to the predicted blur state, thereby improving image quality without hardware changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If hardware improvement (lens design adjustment) is used to remove four-corner blur, then image quality at edge areas is improved, but hardware cost increases and other imaging quality problems (subject shape distortion) become more serious
Solution Approach 1:
The patent replaces hardware improvement (lens design adjustment) with a software-based image processing method. The system uses a Coma prediction module that analyzes image features and position information to predict trailing blur state, then applies adaptive processing algorithms to correct the blur in software, eliminating the need for expensive hardware modifications while avoiding subject shape distortion
Solution Approach 2:
The patent changes the processing parameters dynamically based on predicted blur state. The system adjusts convolution kernel values, processing strength, and algorithm parameters according to the predicted Coma aberration level and direction, enabling precise control over image quality improvement without hardware changes
2Ease of manufacture
If conventional image deblurring techniques are used to remove four-corner blur, then processing is simple, but the trailing blur problem cannot be completely solved because ordinary deblurring does not consider four-corner characteristics
Solution Approach 1:
The patent applies different processing strategies to different regions of the image based on predicted blur characteristics. The Coma prediction module identifies four-corner areas with trailing blur and applies specialized adaptive processing only to those regions, while maintaining simple processing for center areas, thus achieving high precision blur removal where needed without overcomplicating the overall process
Solution Approach 2:
The system dynamically adjusts processing parameters based on real-time Coma prediction. The convolution kernel values, processing strength, and algorithm selection change adaptively according to the predicted trailing blur state, allowing the system to solve the trailing blur problem effectively while maintaining processing simplicity through automated parameter adjustment
Data Source
AI summary
A method performed by an electronic apparatus, includes: obtaining position information of an object in an input image; predicting trailing blur state information of the input image, based on the position information and the input image; and obtaining an output image by performing processing on the input image, based on the predicted trailing blur state information.