Over the past decade structural signal processing is an emerging field, which gained researchers’ intensive attentions in various areas including the applied mathematics, physics, information theory, signal processing, and so on. The structural signal processing is a paradigm of making the revolutionary refresh on theories and methods in the nutshell of traditional signal processing based on the well-known Nyquist-Shannon theory, which will render us a new perspective on the adaptive data acquisition in the task-driven manner. Basically, the structural signal processing includes four research contents (MAMA): (a) Measures for the structural signal, (b) Algorithms for reconstructing the structural signal at the low-complexity computational cost, (c) Methods for smart data acquisition at the low hardware cost and system complexity, and (d) Applications of structural signal processing in applied fields. This paper reviews on the recent progress on the theory and algorithms for structural signal processing, which will provides hopefully useful guide for readers of interest.%结构化信号处理是近年来信息领域发展极为迅猛的一个研究分支，它革新了以Nyquist-Shannon理论为基础的信号处理经典体系的众多结论，开启了面向对象的信息处理的大门，促使挖掘信号的结构性与自适应测量有机结合，推动了信息论、电子学、医疗、应用数学、物理等领域的发展。结构化信号处理研究结构化信号的获取、表征、复原及应用等问题，主要包含4方面内容：(1)研究结构化信号表征与测度的模型和理论；(2)研究结构化信号的复原模型、理论及算法实现；(3)研究信号获取的新体制；(4)研究结构化信号处理的应用。该文以数据和先验两类信息源的融合为主线，讨论了结构化信号处理在信号表征和大尺度信号复原等方面的最新研究结果，并对该领域的发展进行了展望。
展开▼
