Image Denoising Using Gradient-Based Boundary Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image denoising methods blur image boundaries and lose details when strong noise reduction is applied, while weak noise reduction fails to completely remove noise.
Innovation Solution
A method that classifies image pixels as boundary or non-boundary points using gradient values and structure tensors, applying a bilateral filter with different radii to preserve sharpness and remove noise effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If strong denoising is performed on the whole image using a same filter method, then noise is reduced completely, but the boundary of the image is blurred and image details are lost
Solution Approach 1:
The patent applies different filtering strengths to different regions of the image by classifying pixels into boundary and non-boundary regions. Boundary pixels use a first filtering strength that preserves sharpness, while non-boundary pixels use a second filtering strength that removes noise more aggressively. This local differentiation resolves the contradiction by allowing strong denoising in safe regions while protecting important boundary regions from blurring.
Solution Approach 2:
The image is segmented into multiple regions based on gradient magnitude thresholds. Pixels are classified as either boundary pixels (high gradient regions) or non-boundary pixels (low gradient regions). This segmentation enables the application of different filtering strategies to different parts of the image, allowing complete noise removal in non-boundary regions while maintaining boundary sharpness.
2Manufacturing precision
If weak denoising is performed on the whole image using a same filter method, then image boundary sharpness is preserved, but noise cannot be removed completely
Solution Approach 1:
The patent applies different filtering strengths to different regions of the image by classifying pixels into boundary and non-boundary regions. Boundary pixels use a first filtering strength that preserves sharpness, while non-boundary pixels use a second filtering strength that removes noise more aggressively. This local differentiation resolves the contradiction by allowing strong denoising in safe regions while protecting important boundary regions from blurring.
Solution Approach 2:
The image is segmented into multiple regions based on gradient magnitude thresholds. Pixels are classified as either boundary pixels (high gradient regions) or non-boundary pixels (low gradient regions). This segmentation enables the application of different filtering strategies to different parts of the image, allowing complete noise removal in non-boundary regions while maintaining boundary sharpness.
3Manufacturing precision
If different filtering radii are used for boundary and non-boundary points, then both noise removal and boundary sharpness are optimized, but the filtering process becomes more complex
Solution Approach 1:
The image is segmented into multiple regions based on gradient magnitude thresholds. Pixels are classified as either boundary pixels (high gradient regions) or non-boundary pixels (low gradient regions). This segmentation enables the application of different filtering strategies to different parts of the image, allowing complete noise removal in non-boundary regions while maintaining boundary sharpness.
Solution Approach 2:
The patent dynamically changes the filtering radius parameter based on the local image characteristics. Boundary pixels use a first filtering radius that is smaller to preserve detail, while non-boundary pixels use a second filtering radius that is larger to remove noise more effectively. This parameter adaptation allows optimized filtering performance while managing complexity through automated classification.
Data Source
AI summary
A method for implementing image denoising is provided. The method includes: calculating a tangent value of each pixel; determining whether a modulus value of the tangent value of each pixel is less than a preset threshold, if yes, determining a corresponding pixel as a non-boundary point of the image, and performing bilateral filter on a pixel determined as a non-boundary point of the image and pixels which are around the pixel and of which distances to the pixel are less than or equal to a first filtering radius; and if not, determining a corresponding pixel as a boundary point of the image, and performing bilateral filter on a pixel determined as a boundary point of the image and pixels whose distances along tangent directions and opposite directions of the tangent directions to the pixel are less than or equal to a second filtering radius.