The Design And Implemention Of Embedded Image Acquisition Node Based On Compressive Sensing

With the proposal of the concept of the Internet of Things and wisdom of the Earth, the research and application of wireless sensor networks (Wireless Sensor Network WSN) has been an unprecedented development. WSN is composed by a large number of sensor nodes, sensor node deployment through a certain way to establish a wireless network in a certain environment over the completion of the environment and other specific information acquisition and transmission. Sensor nodes usually carry single-chip as the microprocessor, the resulting problem is complex calculations and data storage problems.In recent years Compressive Sensing (CS) by reducing the conventional sampling rate technology to reduce the amount of data in the WSN effectively reduce redundant data acquisition and processing, the completion of the node during the transmission of energy limited become the key technology of WSN research. Through the introduction of embedded technology, with the help of compressive sensing theory, design an image acquisition node. The node is to get rid of a large amount of image data, widely used WSN difficult predicament.This thesis briefly describes related theoretical knowledge of CS, background and significance of the research reviews current status at home and abroad,then investigates the applications of CS in WSN, especially the application on image in detail. Meanwhile design a CS image acquisition node with embedded technology as a development tool. Firstly, the analysis of CS signal sparse representation of the classic methods, use discrete wavelet transforms(DWT) the image signal sparse representation, in order to meet the collected node CS image processing can be performed. Given the design of general use, on the design of the measurement matrix, choices the most commonly used Gaussian random measurement matrix as the measurement matrix of CS. Secondly, set up software and hardware platform on the basis of theoretical research on CS, an image acquisition node based on CS was designed, the algorithm is applied to the node in the form of software;Finally, through on-site testing, observes image restoration results in the case of different sampling rates on acquisition node. Experimental results show that acquisition node can complete the acquisition CS image processing, and sample rate MR=0.7can get a good recovery.