Video Quality Evaluation Using Target-Object HSL Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video quality evaluation methods, whether subjective or objective, struggle to accurately reflect user perception due to reliance on global data rather than focusing on local content, leading to inconsistent and less accurate quality assessments.
Innovation Solution
Evaluating video quality using HSL data from targeted image areas within video frames, aligning with human focus on specific objects, employing image segmentation and data conversion to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If global video data is used for quality evaluation, then the evaluation covers the entire video content, but the accuracy of reflecting user perception deteriorates
Solution Approach 1:
The patent divides the video content into multiple local regions based on object detection, focusing quality evaluation on specific areas containing target objects rather than analyzing the entire video frame. This segmentation approach improves measurement precision by concentrating evaluation resources on perceptually important regions while reducing overall computational complexity.
Solution Approach 2:
The patent applies different evaluation strategies to different local regions of the video. By identifying target objects and their surrounding regions, the system applies quality evaluation metrics specifically to these local areas rather than uniformly across the entire video, thereby improving accuracy where it matters most while reducing unnecessary computations in irrelevant areas.
2Reliability
If HSL data from targeted image areas is used, then the evaluation aligns better with human perception, but the processing complexity increases
Solution Approach 1:
The patent converts video data from standard color spaces to HSL (Hue, Saturation, Lightness) color space parameters. This parameter transformation enables the system to evaluate color quality attributes that better correspond to human visual perception, such as color vibrancy and brightness, while maintaining computational efficiency through targeted application to specific image regions.
3Productivity
If comprehensive video analysis is performed, then all video content is evaluated, but the evaluation efficiency decreases
Solution Approach 1:
The patent extracts and isolates target objects from the video content using object detection algorithms, then focuses quality evaluation exclusively on these extracted regions. This extraction approach improves evaluation efficiency by eliminating processing of background and irrelevant areas, while simultaneously improving accuracy by concentrating analytical resources on perceptually significant content.
Data Source
AI summary
Provided is a method for evaluating video quality, an electronic device and a storage medium, relating to the field of computer technology, and in particular, to the field of artificial intelligence, cloud computing, computer vision and deep learning technologies. The method includes: determining a plurality of video frames containing a target object in a target video; obtaining HSL data of an image area corresponding to the target object in the plurality of video frames; and evaluating a quality of the target video according to the HSL data.