Parallax Image Noise Reduction via Pixel Group Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods are insufficient in reducing noise to achieve high image quality, especially with the advancement of display devices, as they simply combine pixels following Poisson distribution, which is not sufficient for further enhancing image quality.
Innovation Solution
An image processing apparatus and method that calculates the average pixel value of a pixel group from the same region of an object space and shifts these pixels to generate output images with different focus positions, effectively reducing noise by combining parallax images with shifted pixel values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pixels are simply combined following Poisson distribution, then noise is reduced by N-1/2 times, but image quality is insufficient for advanced display devices
Solution Approach 1:
The invention segments pixels into groups based on their imaging relationships. Specifically, it identifies first pixels that image the same object region and second pixels that image different object regions, then processes these segments differently - averaging first pixels for noise reduction while preserving second pixels for detail retention.
Solution Approach 2:
The invention applies different processing qualities to different pixel groups. First pixel groups (imaging same object) receive heavy averaging for noise reduction, while second pixel groups (imaging different objects) receive minimal processing to preserve edge information and structural details, achieving local optimization of image quality.
2Measurement precision
If average pixel value is calculated and substituted for pixel group, then noise is reduced, but pixel alignment may cause artifacts
Solution Approach 1:
The invention dynamically determines pixel combination strategies based on pixel group characteristics. It calculates combination priorities for each pixel group and adjusts the averaging intensity accordingly - using higher averaging for low-priority groups and preserving original values for high-priority groups containing edge or detail information.
Solution Approach 2:
The invention changes the parameter of pixel combination by introducing priority-based weighting. Instead of uniform averaging, it applies different combination intensities based on the calculated priority of each pixel group, thereby adapting the noise reduction strength to the local image content requirements.
Data Source
AI summary
An image processing apparatus capable of generating a plurality of output images having different focus positions by reconstructing an input image, includes a storage unit storing image pickup condition information, and an image processing unit generating the output image from the input image using the image pickup condition information, and the image processing unit obtains the input image that is information of an object space viewed from a plurality of viewpoints that is obtained via an imaging optical system and an image pickup element having a plurality of pixels, calculates an average pixel value of a pixel group of the input image of the same region, and substitutes each pixel value of the pixel group by the average pixel value, and performs combination such that the pixels substituted by the average pixel value are shifted from each other to generate the output image.