摘要
为了提高在线学习资源的压缩存储效率,减少占用云端存储空间,提出基于边缘计算的在线学习资源压缩存储方法.该方法以分布式压缩感知为边缘计算方法,通过构建第一联合稀疏模型JSM-1压缩采集在线学习资源,结合同步正交匹配追踪算法与字典学习算法组成的联合重构算法重构资源,获取稀疏字典原子和测量值,并将其传送至云端服务器,建立云端完备稀疏字典,实现在线学习资源的压缩存储.实验表明,稀疏字典越大且训练学习样本数量越多的资源压缩存储的信噪比越高,资源处理性能越好.用该方法进行压缩在线学习资源后的图片清晰度较高,压缩时间较短,能够节约压缩时间减少存储空间.
In order to improve the compression storage efficiency of online learning resources and reduce the occupation of cloud storage space,a compression storage method of online learning resources based on edge computing is proposed.This method takes distributed compressed sensing as the edge computing method,compresses and collects online learning resources by constructing the first joint sparse model jsm-1,reconstructs resources by combining the joint reconstruction algorithm composed of synchronous orthogonal matching tracking algorithm and dictionary learning algorithm,obtains sparse dictionary atoms and measured values,and transmits them to the cloud server to establish a cloud complete sparse dictionary.Then,compressed storage of online learning resources is implemented.Experiments show that the larger the sparse dictionary and the larger the number of training learning samples,the higher the signal to noise ratio of compressed storage,and the better the resource processing performance;Moreover,the image compressed by this method has high definition and short compression time,which can save a lot of compression time and reduce the storage space occupied by resources.
作者
李媛
LI Yuan(Department of Commerce and electronic information,Tongcheng Teachers College,Tongcheng Anhui 231400)
出处
《宁夏师范学院学报》
2022年第1期76-83,共8页
Journal of Ningxia Normal University
基金
安徽省质量工程项目(2020jxtd264)
安徽省质量工程项目(2020mooc502)
安徽省自然科学重点项目(KJ2020A0892)
安徽省质量工程项目(2020jyxm1996).
关键词
边缘计算
感知算法
联合稀疏模型
云端服务器
稀疏字典
信噪比
Edge computing
Perception algorithm
Joint sparse model
Cloud server
Sparse dictionary
Signal to noise ratio
