Skip Picture Encoding at Video Switching Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-layer or multi-view video encoding, existing technologies face inefficiencies at switching points, leading to redundant coding and increased bitstream size due to encoding both pictures at a switching point, even though only one will be displayed, which complicates both encoding and decoding processes.
Innovation Solution
A method and encoder that encode one picture as a skip picture at the switching point, reducing redundancy by signaling only the necessary information, thereby minimizing the bitstream size and complexity during both encoding and decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If both pictures at the switching point are encoded as regular pictures, then complete video data is available for reference, but coding complexity and bitstream size increase significantly
Solution Approach 1:
The patent extracts only the necessary picture at the switching point for encoding while taking out (excluding) the other picture from the bitstream. By identifying which picture is needed for reference and which can be skipped, the system reduces coding complexity and bitstream size while maintaining the reliability needed for video decoding.
Solution Approach 2:
The patent applies discarding by skipping the encoding of one picture at the switching point when another picture with the same POC value is already available in the bitstream. The discarded picture information can be recovered through reference to the already-encoded picture, maintaining video quality while reducing complexity.
2Reliability
If both pictures at the switching point are encoded as regular pictures, then reference data is available for future predictions, but bitstream size increases
Solution Approach 1:
The patent extracts and encodes only the necessary picture at the switching point while excluding the redundant picture. This ensures that reference data for future predictions is available through the encoded picture, while avoiding the increase in bitstream size that would result from encoding both pictures.
Solution Approach 2:
The patent uses copying by having the decoder copy reference data from the already-encoded picture instead of encoding a duplicate. When one picture at the switching point is skipped, its data can be copied from the other picture that is encoded, maintaining prediction reference availability without increasing bitstream size.
3Adaptability or versatility
If Intra pictures are used for resolution changes, then random access is enabled, but coding efficiency is reduced
Solution Approach 1:
The patent applies dynamics by allowing flexible switching between different picture types (Intra, Inter, skip) at switching points based on the specific coding context. Rather than forcing Intra pictures for all resolution changes, the system dynamically selects the most efficient encoding mode while maintaining random access capability when needed.
Solution Approach 2:
The patent changes the parameter of picture encoding mode at switching points. By using skip pictures with appropriate reference picture selections instead of always using Intra pictures, the system maintains random access capability while improving coding efficiency through more compact representation.
Data Source
AI summary
A multi-layer or multi-view video is encoded by encoding one of a picture in a first layer or view and a picture in a second layer or view coinciding at a switching point defining a switch between the first layer or view and the second layer or view. The other of the picture in the first layer or view and the picture in the second layer or view coinciding at the switching point is encoded as a skip picture. The embodiments thereby reduce complexity of encoding and decoding multi-layer or multi-view video having a switching point and reduce the number of bits required for representing encoded pictures coinciding at the switching point.