Image Processing System Gradient Constraint Artifact Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods, such as those described in Non Patent Document 1, lack constraints on absolute pixel values, leading to artifacts like lost detailed structures, white out, black out, halo effects, and ringing in output images.
Innovation Solution
An image processing system that includes a gradient calculation unit, an indication function calculation unit, a pixel value renewal unit, and a pixel value constraint unit to calculate and constrain pixel values, ensuring they fall within a defined range and approximate the reference image values, thereby producing an output image with improved quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pixel values are renewed to approximate the desired gradient, then the gradient accuracy is improved, but artifacts such as lost detailed structures, white out, black out, halo effect, and ringing occur
Solution Approach 1:
The patent applies preliminary anti-action by introducing a constraint term in the optimization function that prevents pixel values from exceeding valid ranges before artifacts can occur. The constraint term penalizes pixel values that fall outside the valid range [0, 255], thereby preventing white out and black out artifacts before they happen during the gradient approximation process.
Solution Approach 2:
The patent uses an intermediary approach by introducing a constraint function as a mediator between the gradient approximation objective and the pixel value output. This constraint function acts as a bridge that allows gradient optimization while simultaneously enforcing pixel value boundaries, thus preventing artifacts without compromising gradient accuracy.
2Object-generated harmful factors
If pixel values are constrained to fall within a defined range, then artifacts are suppressed, but the gradient approximation accuracy may be reduced
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the optimization objective function to include both gradient approximation terms and constraint terms. The balance between these terms is controlled by parameter weights, allowing the system to maintain gradient accuracy while enforcing pixel value constraints. This enables simultaneous achievement of both goals through parameter optimization.
Solution Approach 2:
The patent uses a composite approach by combining multiple objective functions into a unified optimization framework. The total objective function comprises gradient approximation components and pixel value constraint components, similar to composite materials combining different properties. This composite function allows the system to achieve both gradient accuracy and artifact suppression simultaneously.
3Manufacturing precision
If multiple processing steps are added to constrain and renew pixel values, then image quality is improved, but the processing complexity increases
Solution Approach 1:
The patent applies merging by combining gradient calculation, pixel value renewal, and constraint enforcement into a single unified optimization process. Instead of separate processing steps, all operations are integrated into one optimization framework that simultaneously achieves gradient approximation and pixel value constraint, thereby improving image quality without proportionally increasing processing complexity.
Solution Approach 2:
The patent uses universality by designing a multi-functional optimization framework that performs multiple tasks simultaneously: gradient calculation, pixel value renewal, and constraint enforcement. This universal approach allows a single processing system to handle all image processing functions, reducing overall system complexity compared to separate dedicated modules for each function.
Data Source
AI summary
An image processing system in which, in order to easily analyze input images acquired by sensors, output image quality is improved so that is suitable for a user. It includes: a gradient calculation unit that calculates a desired gradient based on input images; an indication function calculation unit that calculates an indication function for the input images, the indication function defining a range that can be taken by an output image and pixel values of a reference image; a pixel value renewal unit that renews pixel values of one of the input images so as to approximate the desired gradient to produce a renewed image; and a pixel value constraint unit that updates pixel values of the renewed image so as to fall within the range that can be taken by the output image and to approximate the pixel values of the reference image, to thereby obtain the output image.