Adaptive Spatial Image Filter for Noise Reduction and Edge Preservation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Image capture devices, especially in compact multi-purpose mobile devices, face challenges in reducing noise while maintaining image quality due to physical limitations and interference, with existing filtering techniques either blurring images or enhancing noise.

Innovation Solution

An adaptive spatial image filter that adjusts parameter values based on current exposure and scaling levels, employing both horizontal and vertical sharpening filters with configurable low and high pass components to reduce noise and enhance edges, thereby refining image information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If low pass filtering is applied to captured image information, then noise is reduced, but image sharpness deteriorates due to blurring of edges

Engineering Contradiction:
ImprovenoiseVSAvoidimage sharpness
Core Design Contradiction:
Object-affected harmful factorsVSManufacturing precision

Solution Approach 1:

The patent applies different filtering strengths to different spatial regions of the image. The adaptive spatial filter modifies filter coefficients based on local image characteristics such as edge detection and noise levels in specific regions, allowing aggressive noise reduction in uniform areas while preserving edges and fine details where quality is maintained

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The filter dynamically adjusts its parameters based on local image content analysis. The adaptive filter computes different filter coefficients for different regions, transitioning between noise reduction and edge preservation modes depending on the local characteristics, making the filtering process adaptive rather than static

Inventive Principle:
Principle #15Dynamics

2Manufacturing precision

If high pass filtering is applied to captured image data, then sharp edges and contrast are enhanced, but noise is also enhanced

Engineering Contradiction:
Improveedge sharpnessVSAvoidnoise
Core Design Contradiction:
Manufacturing precisionVSObject-affected harmful factors

Solution Approach 1:

The patent applies edge enhancement selectively only in regions where edges are detected, while applying noise suppression in uniform regions. The adaptive filter analyzes local variance and gradient information to determine whether to enhance edges or suppress noise in each region, preventing noise amplification in areas without meaningful edge content

Inventive Principle:
Principle #3Local quality

3Volume of moving object

If image sensor module size is decreased for compact devices, then device compactness is improved, but noise capture increases significantly

Engineering Contradiction:
Improvedevice sizeVSAvoidnoise
Core Design Contradiction:
Volume of moving objectVSObject-affected harmful factors

Solution Approach 1:

The patent changes the parameters of the spatial filter adaptively based on exposure level and scaling factor. By adjusting filter strength, kernel size, and coefficients according to the specific imaging conditions caused by the small sensor size, the system optimizes noise reduction effectiveness for compact device constraints

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2055093B1Adaptive spatial image filter for filtering image information
Publication Date: 2017.11.01 QUALCOMM INC
  • EP2055093B1 patent drawingFigure 1
  • EP2055093B1 patent drawingFigure 2
  • EP2055093B1 patent drawingFigure 3

AI summary

This disclosure describes adaptive filtering techniques to improve the quality of captured image information, such as video or still images. An image sensor captures image information and determines a plurality of parameter values based on a current exposure index and a current scaling factor of the image information. The adaptive spatial image filter includes both horizontal and vertical sharpening filters and configures, i.e., adapts, the horizontal sharpening filter and the vertical sharpening filter based on the plurality of parameter values determined by the image sensor. The adaptive spatial image filter applies the horizontal and vertical sharpening filters to at least one channel of the image information to generate filtered image information.