Image Down-Sampling With Pixel Merging Weights

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image down-sampling methods, such as nearest neighbor interpolation, result in significant loss of image information, leading to lower definition in the down-sampled images.

Innovation Solution

An image processing method that involves acquiring a first image, performing down-sampling to obtain a first down-sampled image, applying average pooling to merge pixel information, determining pixel merging weights, and then merging the down-sampled and pooled images to retain more image information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If nearest neighbor interpolation is used for down-sampling, then the processing speed is fast and the method is simple, but image information is lost and definition is reduced

Engineering Contradiction:

Improvedown-sampling speedVSAvoidimage information loss

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent combines nearest neighbor interpolation with average pooling and pixel merging operations. The down-sampled image is merged with a pooled image generated from the original image, where pixel merging weights are determined based on pixel values and positions. This combination retains more image information while maintaining computational efficiency.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent performs average pooling on the original image before down-sampling to create a pooled image. This preliminary action preserves image information that would otherwise be lost during down-sampling, and the pooled image is then used in the final merging step to recover lost details.

Inventive Principle:

Principle #10Preliminary action

2Device complexity

If nearest neighbor interpolation is used for down-sampling, then the computational complexity is low, but the definition and quality of the down-sampled image deteriorates

Engineering Contradiction:

Improveprocessing complexityVSAvoidimage definition

Core Design Contradiction:

Device complexityVSManufacturing precision

Solution Approach 1:

The patent applies different processing strategies to different regions of the image. Pixel merging weights are determined based on local pixel values and positions, allowing the method to adapt to different image regions. This local adaptation improves image definition while keeping the overall algorithm relatively simple.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent introduces pixel merging weights as a new parameter that combines information from both the down-sampled image and the pooled image. These weights are determined based on pixel values and positions, allowing flexible control over the merging process to improve image quality without excessive complexity.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250069185A1Image processing method, electronic device and storage medium

Publication Date: 2025.02.27 DOUYIN VISION CO LTD

AI summary

An image processing method, an electronic device and a storage medium are provided. The method includes acquiring a first image; performing down-sampling processing on the first image to obtain a first down-sampled image; performing average pooling processing on pixels in the first down-sampled image to obtain a first pooled image; determining a plurality of pixel merging weights corresponding to pixels in the first down-sampled image; and performing pixel merging processing on the first down-sampled image and the first pooled image based on the plurality of pixel merging weights to obtain a second image.