Image Texture Enhancement Without AI Hardware Overhead
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image enhancement methods, such as using Photoshop or AI models, are costly and hardware-intensive, and additional light sources in medical imaging increase device costs and risks, while existing AI methods require high-performance hardware and affect battery life.
Innovation Solution
An image feature enhancement method involving blurring, texture feature extraction, gamma correction, and merging of images to enhance image features, which can be performed on devices with poor hardware and without additional light sources, and is applicable for pre-processing AI models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If AI models are used to calculate and improve image definition, then image quality is improved, but hardware requirements increase and device costs increase
Solution Approach 1:
The patent replaces expensive AI models with a simple, lightweight image processing method that uses basic operations like blurring, difference calculation, and gamma correction. This approach achieves image enhancement without requiring high-performance hardware or complex computational resources, effectively using a simple, disposable algorithm instead of expensive, resource-intensive AI systems
Solution Approach 2:
The patent substitutes complex computational mechanics (AI model calculations) with simpler image processing operations. By using straightforward mathematical operations such as blurring filters, pixel difference calculations, and gamma correction curves, the system replaces the need for complex neural network computations with much lighter mechanical processing steps
2Measurement precision
If AI models are used to improve image definition, then image quality is improved, but battery life is affected due to heavy computational load
Solution Approach 1:
The patent employs a computationally inexpensive image processing algorithm that consumes minimal battery power. By using simple operations like blurring, difference calculation, and gamma correction instead of power-hungry AI models, the method achieves image enhancement while preserving battery life in portable devices
Solution Approach 2:
The patent applies image processing operations selectively and efficiently, performing only the necessary calculations to enhance image definition. By using partial processing (applying enhancements only where needed) and avoiding excessive computational steps, the method reduces energy consumption while still achieving the desired image quality improvement
3Illumination intensity
If additional light sources are used in medical imaging, then image features are displayed, but device costs increase and risk of device damage increases
Solution Approach 1:
The patent extracts and enhances image features through computational processing rather than adding physical light sources. By using image processing techniques to highlight texture features and abnormalities in medical images, the system removes the need for additional NBI light sources, thereby reducing device complexity and cost while maintaining diagnostic image quality
Solution Approach 2:
The patent creates an enhanced copy of the original medical image through computational processing. By generating a processed version of the image with enhanced texture features and abnormalities, the system replicates the effect of additional lighting without physically adding light sources, thus avoiding increased device complexity and risk
Data Source
AI summary
An image feature enhancement method and an electronic device are provided. The image feature enhancement method includes: blurring an original image to obtain a first blurred image; calculating a difference between the original image and the first blurred image to extract a first texture feature image; performing gamma correction on the first blurred image and the first texture feature image respectively to generate a second blurred image and a second texture feature image; and merging the second blurred image and the second texture feature image to generate a final image with enhanced features. The texture feature structure of the image can be quickly enhanced to make the image look clearer.