Black and White Stretch Method for Video Dynamic Range Extension
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for extending the dynamic range of video signals through black and white stretching suffer from limitations such as reliance on linear mapping, which leads to clipping and false contours, and fail to adequately consider mid-levels and pixel distribution, resulting in suboptimal contrast enhancement.
Innovation Solution
A method involving a modified cumulative histogram to determine dynamic range, calculating stretch regions and strengths based on minimum and median levels, using curve structures for mapping, and applying spatial and temporal averaging filters to eliminate false contours and flickering effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If linear mapping is used for black and white stretching, then the dynamic range is extended, but clipping and false contours occur at the darkest and brightest levels
Solution Approach 1:
The patent applies curve structures instead of linear mapping functions to perform black and white stretching. Specifically, cubic splines or polynomial curves are used to map intensity values, which smooths the transitions and eliminates the clipping and false contours that occur with linear mapping at the extreme intensity levels.
Solution Approach 2:
The patent changes the mapping function from linear to non-linear (curved) to resolve the clipping issue. By adjusting the parameters of the curve (such as spline coefficients or polynomial degrees), the system achieves smooth mapping that preserves detail at the darkest and brightest levels while extending dynamic range.
2Illumination intensity
If linear mapping is used with high gain, then contrast enhancement is achieved, but false contours appear in the enhanced frame
Solution Approach 1:
The patent uses curved mapping functions (cubic splines, polynomials) instead of linear functions with high gain. The curves provide smooth, continuous transformations that maintain contrast enhancement while avoiding the sharp transitions and discontinuities that create false contours in the enhanced image.
Solution Approach 2:
The patent implements adaptive stretching where the mapping curve parameters are adjusted based on the input image characteristics (histogram, dynamic range). This dynamic adaptation allows the system to optimize contrast enhancement for each specific image while preventing false contours through smooth, context-aware curve parameters.
3Illumination intensity
If only black stretch is applied, then high contrast feeling is achieved, but dynamic range extension is insufficient
Solution Approach 1:
The patent segments the intensity range into multiple regions (dark, mid, bright) and applies different stretching strategies to each. Black stretch is applied to the dark region to enhance contrast, while white stretch is applied to the bright region to extend dynamic range, with mid-levels handled separately to preserve natural appearance.
Solution Approach 2:
The patent applies different mapping characteristics to different intensity regions. The black stretch uses one curve configuration for the dark region, while the white stretch uses another configuration for the bright region. This local optimization allows simultaneous contrast enhancement and dynamic range extension without compromising either objective.
4Difficulty of detecting and measuring
If thresholds are used to determine pixel frequency, then black stretch strength is detected, but the method fails to consider mid-levels and pixel distribution adequately
Solution Approach 1:
The patent segments the intensity histogram into multiple regions (dark, mid, bright) and analyzes pixel distribution in each region separately. This segmentation allows the system to detect stretch strength for black regions while also considering mid-levels and bright regions, providing comprehensive stretch parameter calculation that accounts for overall pixel distribution.
Solution Approach 2:
The patent moves from a single-threshold approach to a multi-dimensional analysis by examining histogram characteristics across multiple intensity ranges. By considering the distribution of pixels across different intensity levels (not just below a single threshold), the system achieves more accurate stretch parameter detection that reflects the actual image content.
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
Input content for TV receivers commonly has narrow dynamic range; lack of pure dark or lack of pure bright pixels. This situation makes user to feel low contrast for an image on screen. Thus, narrow dynamic range should be extended. For this purpose present invention describes a method for black and white stretch (BWS) of intensity values to increase dynamic range of input video signal. Said method describes a BWS algorithm which has the following steps; detection of the dynamic range of the frame histogram by finding the minimum and maximum existent intensity values, calculation of the median value to determine the starting stretch levels, calculation of the stretch strength according to the histogram distribution for both black and white region dependently and update of the histogram mapping function with the new terms within the stretch boundaries by a specific mapping curve. After above steps, to eliminate sharp increase or decrease, the map is smoothed by a high-order-low-pass filter without changing the end points. Finally, the determined mapping function is weighted with the previous frames' mapping function in order to remove possible flickering effect due to small changes in consecutive frames.