FPAA Accelerator for Machine Vision systems
نویسندگان
چکیده
This article presents a proposition of an FPAA-type programmable accelerator for image preprocessing. The structure of the accelerator is modelled basing on CPLD digital circuits. The innovation here – is using the current mode, which makes it possible to implement the accelerator in nanometre technologies. Another original solution proposed in the work is a reconfigurable multi-output current mirror. The article describes the hardware layer and a method for programming it. An implementation of an RGB-to-YCrCb colour space converter is presented. Moreover physical parameters obtained in post-layout simulations are presented as well. The solution can be used as a standalone programmable circuit or as an IPcore for a larger analogue-digital system. Streszczenie. W artykule przedstawiono propozycję programowalnego akceleratora typu FPAA do wstępnej obróbki obrazu. Struktura akceleratora wzorowana jest na cyfrowych układach CPLD. Innowacyjność polega na wykorzystaniu trybu prądowego, co umożliwia realizację akceleratora w technologiach nanometrowych. Kolejnym oryginalnym rozwiązaniem zaproponowanym w pracy jest rekonfigurowalne wielowyjściowe zwierciadło prądowe. W artykule omówiono warstwę sprzętową oraz metodę jej programowania. Zaprezentowano implementację konwertera przestrzeni barw RGB do YCrCb w akceleratorze i przedstawiono parametry fizyczne uzyskane w symulacjach post-layoutowych. Rozwiązanie może być wykorzystane jako samodzielny układ programowalny lub IP-core większego systemu analogowo-cyfrowego. (Akcelerator FPAA dla systemów wizyjnych).
منابع مشابه
An Embedded Accelerator for Real World Computing
This paper presents a new general purpose accelerator based on the Kress ALU Array III (KrAA-III) a novel field programmable ALU array (FPAA), which efficiently supports high performance arithmetic computations by massive pipelining. The KrAA-III and its underlying concepts will be introduced as a generalization of systolic array structural principles illustrated by a simple real world computin...
متن کاملSoC FPAA Hardware Implementation of a VMM+WTA Embedded Learning Classifier
This paper focuses on the circuit aspects required for an on-chip, on-line SoC large-scale Field Programmable Analog Array (FPAA) learning for Vector-Matrix Multiplier (VMM) + Winner-Take-All (WTA) classifier structure. We start by describing the VMM+WTA classifier structure, and then show techniques required to handle device mismatch. The approach is initially explained using a VMM+WTA as a tw...
متن کاملVMM + WTA Embedded Classifiers Learning Algorithm implementable on SoC FPAA devices
This paper presents a learning algorithm for a VMM + WTA classifier one layer architecture on a Large-Scale Field Programmable Analog Array (FPAA). The technique enables opportunities for embedded, ultra-low power machine learning, techniques typically considered for large servers. To develop this training algorithm, the paper starts by understanding fundamental equivalent transformations for V...
متن کاملModeling, simulation and implementation of circuit elements in an open-source tool set on the FPAA
An open-source simulator to design and implement circuits and systems, replicating the results from the Field Programmable Analog Array (FPAA) is presented here. The fundamental components like the transistors, amplifiers and floating gate devices have been modeled based on the EKV model with minimal parameters. Systems including continuous-time filters and the analog frontend of a speech proce...
متن کاملTwo New Methods of Boundary Correction for Classifying Textural Images
With the growth of technology, supervising systems are increasingly replacing humans in military, transportation, medical, spatial, and other industries. Among these systems are machine vision systems which are based on image processing and analysis. One of the important tasks of image processing is classification of images into desirable categories for the identification of objects or their sp...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2015