RAW Image Combining With Noise-Adaptive Development Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods fail to effectively reduce noise while preserving resolution when combining multiple RAW images, as they either amplify noise or remove high-frequency components, and existing techniques cannot determine noise reduction parameters based on actual noise reduction amounts.
Innovation Solution
An image processing apparatus and method that acquires multiple RAW images, determines noise signal levels before and after combining, and adjusts development parameters based on capturing sensitivity and noise signal levels to reduce noise while maintaining resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If the sensitivity of the image capturing apparatus is increased to obtain an image with appropriate brightness during divided exposure, then the brightness is improved, but the noise of the capturing element is amplified and image quality deteriorates
Solution Approach 1:
The patent divides the total exposure time into multiple short exposure times to capture multiple RAW images. By segmenting the exposure process, the system can use shorter exposure times that generate less noise while still achieving appropriate brightness through combination of multiple images, thereby resolving the contradiction between brightness and noise.
Solution Approach 2:
The patent combines multiple RAW images captured with shorter exposure times to generate a combined RAW image. This merging process allows the system to achieve the brightness equivalent of long exposure while using shorter exposures that produce less noise, thus resolving the brightness-noise contradiction.
2Object-generated harmful factors
If a strong noise reduction parameter is applied to reduce noise in high-sensitivity images, then the noise is reduced, but high-frequency components such as edge portions are removed and resolution is impaired
Solution Approach 1:
The patent performs addition average combine on RAW images before development, and determines the noise reduction parameter based on the actual noise reduction amount achieved by the combine operation. This preliminary action allows the system to customize the noise reduction parameter to match the actual noise level, preventing over-reduction that would remove high-frequency components and impair resolution.
Solution Approach 2:
The patent determines the noise reduction parameter based on feedback from measuring the noise signal level before and after combination. By using the actual noise reduction amount as feedback, the system can adjust the development parameter to achieve appropriate noise reduction without removing important high-frequency components, thus resolving the noise-resolution contradiction.
3Object-generated harmful factors
If addition average combine is performed on developed images, then noise is reduced, but the sense of resolution cannot be restored
Solution Approach 1:
The patent performs addition average combine on RAW images before development rather than on developed images. This preliminary action preserves the high-frequency components and resolution information in the RAW domain, allowing noise reduction without sacrificing resolution, thereby resolving the contradiction between noise reduction and resolution preservation.
4Object-generated harmful factors
If multiple RAW images are combined to reduce noise, then noise is reduced, but the noise reduction amount is not known in advance
Solution Approach 1:
The patent measures the noise signal level before and after the addition average combine operation as a preliminary action. By measuring the actual noise reduction amount that occurs during combination, the system can determine an appropriate development parameter, resolving the uncertainty about noise reduction amount.
Solution Approach 2:
The patent uses feedback from measuring the noise signal level before and after combination to determine the development parameter. This feedback mechanism allows the system to adapt the development parameter to the actual noise reduction achieved, resolving the difficulty of predicting noise reduction amount in advance.
Data Source
AI summary
An image processing apparatus includes a first acquisition unit configured to acquire a plurality of RAW images obtained by continuously capturing a subject, a combining unit configured to combine the plurality of RAW images to generate a combined RAW image, a second acquisition unit configured to acquire first information regarding a noise signal level in any of the plurality of RAW images not combined by the combining unit, a third acquisition unit configured to acquire second information regarding a noise signal level in the combined RAW image after the combine by the combining unit, and a determination unit configured to determine a development parameter for developing the combined RAW image based on capturing sensitivity of a capturing element when the plurality of RAW images are captured and the first information and the second information.