Image Sensor Noise Reduction via Multi-Mode Binning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In low-light environments, electronic devices struggle to capture high-quality images due to decreased light quantity per unit pixel, leading to increased noise and deteriorated image quality, as the resolution of image sensors increases, reducing pixel size and light input.
Innovation Solution
An electronic device method that generates multiple images with different noise characteristics by using an image sensor and processor to combine images, either by converting to a binning mode in low-light conditions or downscaling higher resolution images, and then averaging pixel values to produce a final image with improved resolution and noise characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the resolution of an image sensor is increased, then the image quality is improved, but the size of unit pixel decreases leading to decreased light quantity and increased noise
Solution Approach 1:
The patent combines multiple images with different noise characteristics to produce a final image. By merging a first image (higher resolution, more noise) and a second image (lower resolution, less noise), the system achieves both high resolution and low noise in the output image, resolving the contradiction between resolution and noise
2Measurement precision
If the size of unit pixel is decreased to increase sensor resolution, then more pixels fit in the sensor, but the quantity of accepted light per unit pixel decreases
Solution Approach 1:
The system merges multiple images captured under different conditions (different binning modes) to compensate for the reduced light quantity in high-resolution mode. By combining images from both high-resolution and low-resolution modes, the final image achieves high resolution while maintaining adequate light quantity through the contribution of the lower-resolution image
3Object-affected harmful factors
If multiple images are generated and combined to reduce noise, then noise characteristics are improved, but the processing complexity increases
Solution Approach 1:
The system performs preliminary actions by capturing multiple images with different noise characteristics before final processing. By pre-capturing images in different binning modes and preparing them for combination, the complex noise reduction process is simplified into a structured workflow of capture-prepare-combine, making the overall system more manageable
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method effectively generates images with higher resolution and better noise characteristics in low-light conditions, maintaining image quality by combining images with different noise profiles and adjusting weights based on edge and noise levels.
Implementation Method 1
the electronic device recognizes light inputted through a lens by a sensor, and digitizes and stores the image recognized by the sensor
Data Source
AI summary
An electronic device is provided. The electronic device includes at least one image sensor, and a processor for generating a plurality of images having different noise characteristics through the at least one image sensor, and combining the plurality of images.