Content Adaptive Pixel Intensity Clipping for Video Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital video processing systems face issues with data overflow and underflow due to out-of-range pixel intensity values and intensity shifts, leading to visual artifacts and impairment, especially in dynamic video environments where traditional clipping methods are inadequate.
Innovation Solution
A content-adaptive pixel-intensity clipping method that determines a predefined set of video data, deriving range information from original maximum and minimum values to adaptively clip pixel intensity within a specific range, incorporated into both video encoder and decoder circuits to correct intensity issues during processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional fixed clipping methods are used to prevent data overflow and underflow, then data range compliance is improved, but visual quality deteriorates due to intensity shifts and artifacts in dynamic video environments
Solution Approach 1:
The patent applies dynamics by transitioning from fixed clipping thresholds to adaptive clipping thresholds that dynamically adjust based on local video content characteristics. The clipping range is determined by analyzing statistical properties (mean and standard deviation) of pixel intensities within local regions, allowing the clipping operation to adapt to varying content dynamics across different frames and spatial locations, thereby preventing visual artifacts while maintaining data range compliance
Solution Approach 2:
The patent implements local quality by performing clipping operations on a local basis rather than globally. Different regions of the video frame are processed with region-specific clipping thresholds derived from local statistical analysis. This allows each local region to have optimized clipping parameters matched to its specific content characteristics, improving visual quality while ensuring data compliance
2Object-affected harmful factors
If digital filters are applied to process video data for noise suppression and edge sharpening, then video quality is improved, but data overflow and underflow occur causing intensity inversion artifacts
Solution Approach 1:
The patent applies preliminary action by performing clipping operations immediately after digital filtering processes. The clipping step serves as a preventive measure that corrects intensity values before they cause visual artifacts. By calculating appropriate clipping thresholds based on local statistics and applying them right after filtering, the system proactively prevents data overflow and underflow issues before they manifest as intensity inversion artifacts
3Productivity
If quantization and rounding operations are performed during video coding, then compression efficiency is improved, but intensity shift occurs causing visual impairment
Solution Approach 1:
The patent implements feedback by using statistical information (mean and standard deviation) calculated from the processed video data to dynamically adjust clipping thresholds. This feedback mechanism allows the system to adapt to intensity shifts caused by quantization and rounding operations, compensating for their effects and maintaining visual quality while preserving compression efficiency
Data Source
AI summary
A system and method of content adaptive pixel intensity processing are described. The method includes receiving a predefined set of processed video data configured from the processed video data, deriving a range information associated with an original maximum value and an original minimum value for a predefined set of original video data, wherein the predefined set of processed video data is derived from the predefined set of original video data, and adaptively clipping pixel intensity of the predefined set of processed video data to a range deriving from the range information, wherein the range information is incorporated in a bitstream and represented in a form of the original maximum value and the original minimum value, prediction values associated with a reference maximum value and a reference minimum value, or a range index associated with a predefined range set.