Video Quality Assessment Network Using Block Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for evaluating video quality are inefficient due to the need for lengthy training of human visual system models and the difficulty in modeling subjective quality perception, which hinders effective encoding and decoding of high-resolution images.
Innovation Solution
A method involving a video quality assessment network that splits images into blocks, determines sensitivity and quality assessment information, and combines these to calculate a final image quality score, using deep learning to reflect subjective quality perception in image encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a human visual system model is used to evaluate video quality, then measurement precision of subjective quality is improved, but training time and computational complexity increase excessively
Solution Approach 1:
The patent divides the video evaluation into multiple frames and processes each frame independently through the neural network. The quality assessment is performed on individual frames rather than requiring comprehensive training on entire video sequences, significantly reducing training time while maintaining assessment accuracy.
Solution Approach 2:
The patent uses a pre-trained neural network model that has been trained on a large dataset of video quality assessments. Instead of training a new model from scratch, the system copies and applies the pre-trained model to evaluate video quality, eliminating the need for excessive training time while preserving measurement precision.
2Measurement precision
If comprehensive video quality assessment is performed, then measurement precision is improved, but computational complexity and processing time increase
Solution Approach 1:
The patent segments the video into individual frames and processes each frame through separate neural network pathways (full reference, reduced reference, and no reference quality assessment). This segmentation allows for comprehensive quality evaluation without requiring a single overly complex monolithic system.
Solution Approach 2:
The patent implements a multi-functional assessment system that can perform multiple types of quality assessments (full reference, reduced reference, and no reference) using a unified neural network framework. This universal approach maintains measurement precision across different assessment scenarios while avoiding the need for separate specialized systems for each assessment type.
3Measurement precision
If detailed block-based quality assessment is performed, then measurement precision is improved, but processing time increases
Solution Approach 1:
The patent processes video frames in a periodic manner, evaluating quality at specific frame intervals rather than continuously analyzing every frame. This periodic assessment maintains measurement precision for perceived quality while significantly improving processing speed and productivity by reducing the total number of assessments required.
Solution Approach 2:
The patent divides each video frame into multiple blocks and performs quality assessment on individual blocks. This segmentation allows for parallel processing of different blocks, maintaining detailed measurement precision while improving overall processing throughput through concurrent computation.
Data Source
AI summary
Proposed are a method and apparatus for evaluating the quality of an image, the method including obtaining blocks each having a predetermined size by splitting a target image for evaluating a quality and a reference image that is to be compared with the target image, determining sensitivity information and quality assessment information of each of the blocks by inputting the blocks to a video quality assessment network, and determining a final image quality assessment score of the target image by combining the pieces of quality assessment information of the blocks with each other, based on the pieces of sensitivity information of the blocks.