Deep Learning Noise Reduction Parameter Transition Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise reduction technologies using deep learning (DLNR) often cause shocks in images when parameters are switched, and interpolation methods generate images that differ from reality, leading to abrupt transitions and poor image quality.
Innovation Solution
An image processing apparatus and method that employs a trained model with multiple parameters to smoothly adjust noise reduction settings across multiple images, using a control amount and reference data to minimize shocks during parameter changes, ensuring a high-quality, smooth noise reduction effect.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If deep learning noise reduction parameters are switched to adapt to different scenes, then noise reduction effectiveness is improved, but image shock occurs during parameter transitions
Solution Approach 1:
The patent implements dynamic parameter adjustment by transitioning DLNR parameters gradually across multiple frames rather than switching abruptly. The parameter transition amount is calculated based on frame differences, enabling the system to adapt to different scenes while maintaining image stability during transitions.
Solution Approach 2:
The patent introduces an intermediate transition mechanism that mediates between different DLNR parameter sets. By calculating parameter transition amounts and applying gradual changes across multiple frames, it creates a smooth bridge between parameter states, preventing direct shocks while maintaining adaptability.
2Manufacturing precision
If noise reduction strength is increased to improve image quality in ultra-low light, then noise is reduced more effectively, but processing complexity increases
Solution Approach 1:
The patent dynamically adjusts DLNR parameters based on scene characteristics and frame differences. By changing parameters adaptively rather than using fixed high-strength processing, it achieves effective noise reduction while optimizing processing complexity according to actual needs.
3Adaptability or versatility
If parameter transitions are made abruptly to respond quickly to scene changes, then adaptability is improved, but image shock occurs
Solution Approach 1:
The system implements dynamic parameter transitions that adapt to scene changes while maintaining continuity. By calculating appropriate transition amounts based on frame differences and applying changes gradually across multiple frames, it achieves both scene adaptability and image continuity.
Solution Approach 2:
The patent ensures continuous parameter adjustment across multiple frames rather than abrupt changes. The useful action of noise reduction continues smoothly with gradually changing parameters, maintaining image continuity while adapting to scene transitions.
Data Source
AI summary
An image processing apparatus comprising: one or more memories storing instructions; and one or more processors executing the instructions to: execute image processing of an obtained input image via a trained model including a first parameter obtained by training on an image noise characteristic, reduce noise of an input image subjected to the image processing via a second parameter, combine the input image and an input image with the noise reduced on a basis of a third parameter, and in a case of changing one of the first parameter, the second parameter, and the third parameter, change and set a parameter to be changed on a basis of one of a control amount and preset reference data across a plurality of input images successively obtained.