| Popis: |
The significant consumption of video data over limited bandwidth mostly compromises in video quality. Video quality assessment (VQA) algorithms aim to estimate the quality of a distorted video data. For this, the scores are created in such a way that agrees with the quality judgments aligned to human visual perception. To measure the quality of the video, we need to compare the subjective score and objective score. Video quality assessment (VQA) is a challenging task due to the complexity of modelling perceived quality characteristics in both spatial and temporal domains. We proposed novel work on feature-based VQA. SIFT and SURF are used to compare the features in the original video against features of the distorted video. A mechanism using weighted features is illustrated to provide a better quality assessment. The subjective score is available in many image databases. Ex. LIVE, IPVL, and CSIQ database. An objective score is obtained by using formulas, and it is used to find a correlation between subjective score and objective score. The efficacy of the proposed method is evaluated against state-of-the-art VQA algorithms. Our method is observed to be consistent with the best VQA results. |
