Adaptive Image Processing for Mixed JPEG and HEIF Value Ranges
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods using machine learning models struggle with precision when handling images in different formats such as JPEG and HEIF, due to differences in representable value ranges and image capturing modes, leading to suboptimal results when applying models trained on one format to images in another.
Innovation Solution
An image processing method that adjusts normalization and denormalization processes based on the specific image format and capturing mode, using a machine learning model trained on images in the same format to enhance precision, particularly through techniques like upscaling, blur correction, and contrast enhancement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a machine learning model is trained on images in one format (e.g., JPEG), then the processing speed and simplicity are improved, but the manufacturing precision deteriorates when processing images in different formats (e.g., HEIF) due to differences in representable value ranges
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting normalization parameters based on the input image format. The system detects the image format (JPEG, HEIF, etc.) and selects appropriate normalization parameters tailored to each format's representable value range. This allows a single machine learning model to process multiple image formats with high precision without requiring separate models for each format, thus maintaining both processing speed and manufacturing precision.
2Device complexity
If a single machine learning model is used for all image formats, then the device complexity is reduced, but the measurement precision deteriorates due to differences in image capturing modes and representable value ranges
Solution Approach 1:
The patent implements dynamics by making the normalization process adaptive rather than static. The system dynamically adjusts normalization parameters based on the detected image format and capturing mode. This dynamic adaptation allows a single machine learning model to achieve measurement precision comparable to format-specific models, resolving the contradiction between device complexity and measurement precision.
3Manufacturing precision
If normalization parameters are fixed for a specific image format, then the manufacturing precision is improved for that format, but the adaptability deteriorates when processing images in other formats
Solution Approach 1:
The patent achieves universality by creating a single machine learning model with adaptive normalization capabilities that can process multiple image formats (JPEG, HEIF, and others) with different representable value ranges. The system detects the input format and automatically selects appropriate normalization parameters, enabling one model to perform the function of multiple format-specific models. This resolves the contradiction by maintaining high manufacturing precision across all supported formats while maximizing adaptability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method and the like for performing image processing with higher precision on various image data using a machine learning model are provided. The method includes obtaining an input image and range information about pixel values of the input image, selecting at least one machine learning model from among a plurality of machine learning models based on the range information, and generating an estimated image by inputting the input image to the selected machine learning model. Alternatively, the method includes obtaining an input image and range information about the input image, and generating an estimated image by inputting the input image and the range information to a machine learning model.