Video Signal Processing Luminance Correction Using Previous Frame Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video signal processing methods struggle to achieve smooth transitions in luminance levels between frames, especially during scene changes, as they correct luminance on a per-frame basis without considering previous frame data.
Innovation Solution
A video signal processing method that divides frames into rectangular blocks, calculates feature quantities for each block, and uses correction data based on both current and previous frame data to adjust luminance, ensuring smoother transitions by incorporating temporal luminance information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If luminance correction is performed on a per-frame basis using conventional methods, then the processing complexity is low and the calculation is simple, but the luminance transition between frames becomes discontinuous and unnatural during scene changes
Solution Approach 1:
The patent divides each frame into multiple blocks (e.g., 8x8 or 16x16 pixel blocks) and processes luminance correction independently for each block. This segmentation allows the system to apply different correction strategies to different regions, improving overall luminance transition smoothness while keeping individual block processing computationally manageable
Solution Approach 2:
The patent performs preliminary luminance correction on a previous frame before processing the current frame. By pre-calculating correction values for the previous frame and storing them, the system reduces the computational burden during current frame processing while maintaining smooth luminance transitions across frame boundaries
2Reliability
If luminance correction uses only current frame data, then the processing speed is fast and real-time performance is achieved, but the transition between frames becomes abrupt especially during scene changes
Solution Approach 1:
The system performs luminance correction calculations for the previous frame in advance, storing the correction results for later use. This preliminary processing ensures that when the current frame is being displayed, the correction data is already prepared, maintaining real-time performance while achieving smooth transitions
Solution Approach 2:
The patent maintains continuous luminance correction across frame boundaries by incorporating data from both the previous and current frames. The correction process operates continuously on overlapping frame data, ensuring uninterrupted smooth transitions without sacrificing processing speed
3Reliability
If block-based luminance correction incorporating previous frame data is implemented, then smoother luminance transitions are achieved, but the calculation complexity and processing load increase
Solution Approach 1:
By dividing frames into blocks, the patent enables parallel processing of multiple blocks simultaneously. This segmentation transforms a single complex calculation into multiple simpler parallel calculations, improving processing efficiency while maintaining smooth luminance transitions across the entire frame
Solution Approach 2:
The patent applies different luminance correction characteristics to different blocks based on their local properties. Blocks experiencing large luminance changes receive different correction treatment than blocks with stable luminance, optimizing processing efficiency by focusing computational resources where they are most needed
Data Source
AI summary
A video signal processing method includes: dividing a current frame which is to be processed into a plurality of blocks which are predetermined rectangular blocks; calculating, for each of the plurality of blocks divided into in the dividing, a feature quantity relating to luminance of pixels in the block; calculating first correction data, based on a feature quantity calculated from each of first blocks located above a first horizontal boundary in the current frame and a feature quantity calculated from each of one or more blocks located below a first horizontal boundary in a previous frame which has been processed immediately before the current frame; and correcting luminance of pixels in each of the first blocks in the current frame, based on the first correction data calculated in the calculating.