摘要
针对棉花视频通信网络中质量评价方法忽略本身动态变化特性的问题,提出一种基于图像帧序列显著变化特性的加权视频质量评价方法。该方法从单帧视频的亮度、色度、对比度3个方面提取图像特征,根据灰度共生矩阵求取图像纹理的角二阶矩,并以角二阶矩作为衡量视频动态变化程度的权重,最后通过视频帧加权的方式来得到整个视频的质量评价结果。针对LIVE数据集上的实验测试结果表明,与典型视频质量评价模型相比,该方法评价结果与主观结果之间的性线性相关系数达到0.921,离出率为0.408,能够准确反映人眼视觉感知,评价结果更接近于主观评价,有利于建立适用于棉花监控系统的通信质量评价体系。
Aiming at the problem that the quality evaluation method in cotton video communication network ignores its dynamic change characteristics,a weighted video quality evaluation method based on the significant change characteristics of image frame sequences is proposed. The method extracts image features from three aspects of brightness,chromaticity and contrast of single-frame video,and obtains the second-order moment of the image texture according to the gray-scale co-occurrence matrix,and uses the second-order moment of the angle as the weight to measure the dynamic change degree of the video. Finally,the quality evaluation result of the entire video sequence is obtained by means of video frame weighting. The experimental results on the LIVE dataset show that compared with the typical video quality evaluation model,the linear correlation coefficient between the evaluation result and the subjective result is 0.921,and the exit rate is 0.408,which can accurately reflect the human visual perception. The evaluation results are closer to subjective evaluation,which is conducive to the establishment of a communication quality evaluation system suitable for cotton monitoring systems.
作者
齐英兰
QI Ying lan(College of Light Industry Vocational,Zhengzhou University of Light Industry,Zhengzhou 450002,China)
出处
《电子设计工程》
2020年第4期109-113,共5页
Electronic Design Engineering
基金
国家重点研发计划项目资助(2018YFD0700403)。
关键词
帧序列
显著变化
角二阶矩
视频通信网络
质量评价
video frame
frame significantly change
energy
video communication network
quality evaluation
