研究目的
Investigating the effect of region of interest (ROI) on the overall image quality assessment using subjective rating and objective psychophysiology measurement methods.
研究成果
The ROI compressed stimuli had earlier and higher P300 components than NROI compressed stimuli, indicating it's easier to detect quality differences in ROI compressed images. The study demonstrates the feasibility of using ERPs technology for image quality assessment.
研究不足
The study was limited to a specific set of images and compression techniques. The sample size was small, and the participants were not professionals in the field of quality of experience (QoE).
1：Experimental Design and Method Selection:
The study used subjective rating and objective psychophysiology measurement methods (ERPs) to assess image quality. Different image quality levels were created by compressing only the ROI or background area with different σ values of Gaussian Kernel Function.
2：Sample Selection and Data Sources:
Twelve students from Southeast University participated. Three test images were selected from the Middlebury stereo database and other free online images.
3：List of Experimental Equipment and Materials:
A 55-inch SONY BRAVIA LCD TV, SMI Remote Eye-tracking Device (RED), 64-channel Neuroscan SynAmps2 amplifier, and a 64-channel QuickCap EEG system were used.
4：Experimental Procedures and Operational Workflow:
The experiment had two phases: subjective rating of image quality and discrimination of differences between images while recording EEG signals.
5：Data Analysis Methods:
EEG data were analyzed for P300 amplitude and latency. Subjective scores and detection rates were analyzed using ANOVA and Pearson linear correlation coefficient.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
