Flash Picture Encoding Using Artificial Non-Flash Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies, such as MPEG-2, H.263, MPEG-4 AVC/H.264, and VC-1, face significant bit rate fluctuations and increased bit usage when encoding flash pictures due to abrupt intensity changes, which are not effectively managed by current methods like motion estimation and weighted prediction.
Innovation Solution
The method involves encoding an artificial non-flash or de-flashed picture instead of the real flash picture, with additional parameters to model the flash effect, allowing the decoder to regenerate the flash picture, thereby reducing the number of bits required for encoding and avoiding bit rate bursts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion estimation and intra coding are used to encode flash pictures, then picture quality is maintained, but bit rate increases significantly causing bit rate fluctuation
Solution Approach 1:
The patent extracts the flash effect as a separate component from the picture content. Instead of encoding the entire flash picture with high bit rate, the method separates the flash effect parameters (intensity, position, duration) from the underlying picture data, encoding only the essential components to maintain quality while reducing overall bit rate.
Solution Approach 2:
The patent changes the encoding parameters by using differential coding for flash pictures. It encodes the difference between flash and non-flash states rather than the absolute values, and uses selective intra/inter coding modes based on flash detection, thereby reducing the quantity of encoded data while preserving perceptual quality.
2Quantity of substance
If weighted prediction is used to reduce bit rate of flash pictures, then bit rate decreases, but coding efficiency is insufficient for non-uniform intensity changes
Solution Approach 1:
The patent applies local quality by using different coding strategies for different regions of the picture. Flash-affected regions use differential coding with flash effect parameters, while non-flash regions use standard prediction methods. This localized approach improves coding efficiency for the specific problem of non-uniform intensity changes caused by flashes.
Solution Approach 2:
The patent segments the picture coding process into flash detection, flash effect parameter extraction, and differential encoding stages. It also divides the picture into macroblocks and applies selective coding modes (intra vs inter) based on flash presence in each region, improving overall coding efficiency for flash-containing sequences.
3Manufacturing precision
If multiple reference frames are used to encode non-flash pictures, then coding efficiency improves, but flash picture encoding still produces excessive bits
Solution Approach 1:
The patent performs preliminary flash detection and flash effect parameter extraction before the actual picture encoding. By identifying flash pictures in advance and preparing differential coding parameters beforehand, the method enables efficient encoding of flash pictures without requiring excessive bits, while still allowing non-flash pictures to benefit from multiple reference frame prediction.
Data Source
AI summary
When video sequences are obtained from e.g. news, interviews, conferences and sports matches, flashlight often appears in the video due to photographing. When encoding the video sequence, the intensity changes a lot so that the motion estimation can not find a well-matching block in a previous picture. Accordingly, the video encoder will encode in intra mode and thereby generate much more bits than for the neighboring non-flash pictures. According to the invention, not the real flash picture is encoded in the video sequence but an artificial non-flash or de-flashed picture, and some indications and parameters are inserted into the bit stream to model the flash effect. The decoder can regenerate the flash picture by reconstructing the non-flash picture and then adding the flash effect. The number of bits required for coding flash pictures is thereby significantly reduced.