Image Processor Random Noise Superimposition for False Contour Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-definition liquid crystal display panels experience image quality degradation due to steplike increases and decreases in gray levels, leading to visible errors and false contours when converting 16-bit gray level images to 8-bit representations.
Innovation Solution
An image processor is developed with a random number sequence generation section, a random number-superimposed luminance variable generation section, and a random number-superimposed image signal generation section, which superimpose a random number sequence on the luminance variable and image signal to distribute errors and minimize false contours and waving.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If 16-bit gray level image is converted to 8-bit gray level for display, then the image can be displayed on 8-bit liquid crystal panels, but steplike errors and false contours become visible due to gray level quantization
Solution Approach 1:
The patent applies dithering processing before the gray level conversion to preliminarily distribute the quantization errors. By adding noise to the image signal prior to conversion, the errors that will occur during 16-bit to 8-bit conversion are distributed across multiple pixels, preventing visible steplike artifacts and false contours in the final displayed image
Solution Approach 2:
The patent changes the statistical parameters of the image signal by superimposing a random noise signal with specific characteristics (standard deviation, mean value) onto the original image. This parameter modification transforms the deterministic quantization errors into stochastic distributed errors, improving perceived image quality after conversion
Data Source
AI summary
Disclosed herein is an image processor including: a random number sequence generation section adapted to generate a random number sequence; a random number-superimposed luminance variable generation section adapted to generate a random number-superimposed luminance variable by superimposing the random number sequence on a luminance variable; and a random number-superimposed image signal generation section adapted to generate a random number-superimposed image signal by superimposing the random number-superimposed luminance variable on an image signal.