Parca Giorgia, Teixeira Pedro, Teixeira Antonio
Instituto de Telecomunicações, Campus Universitartio de Santiago, Aveiro, Portugal.
Appl Opt. 2013 Apr 20;52(12):2932-9. doi: 10.1364/AO.52.002932.
Fast data processing and compression methods based on wavelet transform are fundamental tools in the area of real-time 2D data/image analysis, enabling high definition applications and redundant data reduction. The need for information processing at high data rates motivates the efforts on exploiting the speed and the parallelism of the light for data analysis and compression. Among several schemes for optical wavelet transform implementation, the Haar transform offers simple design and fast computation, plus it can be easily implemented by optical planar interferometry. We present an all optical scheme based on an asymmetric couplers network for achieving fast image processing and compression in the optical domain. The implementation of Haar wavelet transform through a 3D passive structure is supported by theoretical formulation and simulations results. Asymmetrical coupler 3D network design and optimization are reported and Haar wavelet transform, including compression, was achieved, thus demonstrating the feasibility of our approach.
基于小波变换的快速数据处理和压缩方法是实时二维数据/图像分析领域的基本工具，可实现高清应用并减少冗余数据。对高数据速率信息处理的需求推动了人们努力利用光的速度和并行性进行数据分析和压缩。在几种实现光学小波变换的方案中，哈尔变换设计简单、计算快速，并且可以通过光学平面干涉测量法轻松实现。我们提出了一种基于非对称耦合器网络的全光方案，用于在光域中实现快速图像处理和压缩。通过理论公式和仿真结果支持了通过三维无源结构实现哈尔小波变换。报告了非对称耦合器三维网络的设计和优化，并实现了包括压缩在内的哈尔小波变换，从而证明了我们方法的可行性。
