Image Sensor Pixel Brightness Correction via Dual Conversion Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image sensors with shared pixel structures face issues with varying saturation levels among pixels, leading to artifacts in bright areas, as one pixel may saturate before others, affecting the dynamic range and image quality.
Innovation Solution
An image processing system that includes a pixel brightness measurement engine, comparator, selector, and interpolator to measure, compare, and correct pixel brightness by selecting or interpolating values based on saturation thresholds, ensuring consistent output across pixels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a shared pixel structure is used to decrease pixel area, then pixel area is reduced, but saturation level varies between pixels causing artifacts
Solution Approach 1:
The patent creates a copy of the pixel signal through dual conversion paths: one path converts pixel signals to grayscale values, and another path converts the same pixel signals to luminance values. This copying mechanism allows the system to have redundant representations of the same data, enabling artifact reduction by comparing and selecting between the grayscale path (which may have artifacts) and the luminance path (which serves as a reference for correction).
Solution Approach 2:
The patent transforms the pixel signal into two different parameter representations: grayscale values and luminance values. By changing the parameter space in which the signal is represented, the system can identify and correct artifacts. The luminance values serve as a reference parameter that helps detect and reduce artifacts in the grayscale path, thereby improving saturation level consistency across pixels.
2Measurement precision
If pixels are measured individually for brightness correction, then correction accuracy is improved, but processing time increases
Solution Approach 1:
The patent merges the brightness correction process for multiple pixels by using the luminance path as a shared reference for all pixel corrections. Instead of independently processing each pixel through separate correction algorithms, the system converts all pixel signals to luminance values once and uses this common luminance representation to guide the artifact reduction process across all pixels, thereby reducing overall processing time while maintaining correction accuracy.
Solution Approach 2:
The patent performs preliminary conversion of pixel signals to both grayscale and luminance values before the actual artifact reduction process. By pre-computing these transformed representations, the system prepares the data in advance for efficient comparison and correction. This preliminary action allows the subsequent artifact reduction step to operate more quickly by working with pre-processed data rather than performing transformations during the correction phase.
Data Source
Figure 1
Figure 2A~2B
Figure 3
AI summary
An image processor is provided for correcting brightness of saturated pixels of a captured image. The image processor can include a pixel saturation determiner that whether one or more pixels in an image sensor have been saturated by comparing pixel brightness levels of the pixels to a predetermined saturation threshold. Moreover, the image processor includes an image enhancer that generates a corrected image without artifacts due to the saturated pixel(s) by replacing the pixel brightness of the saturated pixel(s) with a pixel correction value that is based on a pixel brightness of one or more unsaturated pixel in the image sensor.