Depth Image Intra-Scene Prediction for Multiview Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for encoding depth images in multiview video encoding struggle with efficiently reducing redundancy and minimizing errors, particularly at object boundaries where depth values change significantly, leading to increased complexity in encoding and decoding processes.
Innovation Solution
An inter-layer video decoding method that obtains prediction mode information for a current block, generates a prediction block, predicts an average value using specific pixel values, and decodes the depth image using this average value, thereby reducing the complexity of the encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a calculation process for summing all pixel values belonging to a prediction block and dividing the sum by the number of corresponding pixels is used to calculate a DC value, then the DC value is accurate, but the complexity of encoding and decoding apparatus increases
Solution Approach 1:
The patent extracts only the necessary pixel values (corner pixels and/or center pixel) from the prediction block to calculate the DC value, rather than using all pixel values. This selective extraction maintains sufficient accuracy while significantly reducing computational complexity in both encoding and decoding apparatus.
Solution Approach 2:
The patent applies partial action by using only a subset of pixel values (specifically corner pixels and/or center pixel) instead of the complete set of all pixel values. This partial sampling approach provides a practical balance between calculation accuracy and computational efficiency, avoiding the excessive complexity of processing every pixel.
2Adaptability or versatility
If depth image encoding is performed to synthesize intermediate viewpoint images, then more viewpoints can be generated, but errors at object boundaries with significantly changing depth values increase
Solution Approach 1:
The patent applies local quality by treating different regions of the depth image differently. Specifically, it pays special attention to object boundaries where depth values change significantly, using appropriate prediction modes for these critical regions to minimize errors, while allowing more flexibility in regions with less significant depth variations.
Solution Approach 2:
The patent changes parameters by using different prediction modes (DC mode, planar mode, DMM mode) for different blocks depending on their characteristics. By adapting the prediction mode to the local depth variation characteristics, the patent maintains high synthesized image quality while enabling versatile viewpoint generation.
3Productivity
If multiview video encoding is performed to compress basic viewpoint and expanded viewpoints, then compression efficiency is improved, but data redundancy between time points and viewpoints increases
Solution Approach 1:
The patent segments the depth image into multiple blocks and applies different prediction modes to different blocks based on their characteristics. This segmentation allows for more efficient compression by adapting to local variations, thereby reducing overall data redundancy while maintaining compression efficiency.
Solution Approach 2:
The patent changes encoding parameters by selecting different prediction modes (DC, planar, DMM) for different blocks. This parameter adaptation optimizes the balance between compression efficiency and redundancy reduction, allowing the system to achieve better compression ratios while minimizing unnecessary data repetition.
Data Source
AI summary
An inter-layer video decoding method according to an embodiment of the present invention may comprise the steps of: obtaining prediction mode information for a current block of a depth image from a bitstream; creating a prediction block of the current block on the basis of the obtained prediction mode information; predicting an average value for the prediction block by using the pixel values of a designated location of the prediction block; and decoding the depth image by using the average value for the prediction block.