立体图像质量主观评价被试数量及数据的确定
首发时间:2014-07-16
摘要:立体图像质量主观评价是衡量立体图像质量客观评价方法是否准确的重要依据,而主观实验中的诸多不稳定因素影响了实验结果的精确度。因此,为了得到更加可靠的实验结果,本文中提出了一种主观评价中确定被试者数量和筛选实验数据的方法。通过分析预处理数据,逆用置信区间公式确定主观评价中被试者的数量;根据图像质量主观评价标准的规定和格鲁布斯检验法,对实验数据进行筛选,剔除异常值。实验结果表明,被试者数量确定算法和实验数据筛选方法切实可行,既保证了实验结果的可靠性,又降低了实验复杂度,节约了实验成本。同时,从实验结果还可以得出不同比例的压缩失真对立体图像质量主观评价均值MOS(Mean Opinion Score)的影响程度,以及被试者数量的确定与立体图像本身质量的关系。?????
关键词: 计算机技术 主观评价 立体图像 被试者数量 数据筛选
For information in English, please click here
Determining the Number of Observers and Selecting the Experimental Data in Subjective Quality Assessment Method of Stereo Image
Abstract:Subjective quality assessment of stereo image plays an important role to measure the accuracy of the objective evaluation methods of stereo image. To get more accurate result, we propose a new method to determine the number of observers and select the experimental data in subjective assessment. The number of observers is determined though preprocessing the data and reversing confidence interval formula; the experimental data is selected according to the standards of subjective image quality evaluation and the Grubbs test. The experimental result demonstrates that the proposed method is feasible, the algorithms not only ensure the reliability of experimental result but also reduce the experimental complexity and save the experiment cost. Meanwhile,from the experimental result we can know the influence between the compression and the MOS(Mean Opinion Score),as well as the relationship between the quality of images and the number of observers.
Keywords: Computer technology, Subjective assessment, Stereo image, Observers number, Data selecting
论文图表:
引用
No.4603841989633140****
同行评议
勘误表
立体图像质量主观评价被试数量及数据的确定
评论
全部评论0/1000