A hardware-software architecture for computer vision systems

  title={A hardware-software architecture for computer vision systems},
  author={A. O. Fernandes and L. F. E. Moreira and Glauber Tadeu S. Carmo and J. M. Mata},
  journal={Journal of Engineering and Technology for Industrial Applications},
Computer vision systems are based on capture and processing of images, from which information for the application is extracted. This paper presents a hybrid platform which can be easily configured and efficiently used for a large number of applications in computer vision. It consists of a dedicated processor integrated to the capture element, connected to a general-purpose computer for the highlevel application processing. The camera module consists of a DSP executing a middleware which… Expand


Computer Vision Algorithms on Reconfigurable Logic Arrays
The use of the custom computing approach to meet the computation and communication needs of computer vision algorithms by customizing hardware architecture at the instruction level for every application so that the optimal grain size needed for the problem at hand and the instruction granularity can be matched. Expand
NETRA: A Hierarchical and Partitionable Architecture for Computer Vision Systems
A multiprocessor architecture, called "NETRA," for computer vision systems, which is a hierarchical architecture with a tree-type control hierarchy, and a qualitative evaluation and preliminary performance results of a cluster of NETRA. Expand
High-performance computing for vision
The main focus of the paper is on effectively using commercial-off-the-shelf (COTS) based general purpose parallel computing platforms to realize high speed implementations of vision tasks. Due toExpand
The DARPA Image Understanding Benchmark for Parallel Computers
A new effort to evaluate parallel architectures applied to knowledge-based machine vision is described, which can be used to gain insight into processor strengths and weaknesses and may help to guide the development of the next generation of parallel vision architectures. Expand
Performance Benchmark of DSP and FPGA Implementations of Low-Level Vision Algorithms
This paper reports on a performance benchmark of different implementations of some low-level vision algorithms implemented on both Digital Signal Processor and Field Programmable Gate Array high-speed embedded platforms. Expand
Architectural requirements of image understanding with respect to parallel processing
An overview of the architectural requirements for parallel processing in support of real-time, knowledge-based computer vision is given, concluding that traditional architectural approaches to parallel processors are suboptimal. Expand
Machine vision - theory, algorithms, practicalities
This work focuses on low-level processing on the three-dimensional world tackling the pespective n-point problem motion invariants and their applications and the need for speed - real-time electronic hardware systems. Expand
Smart Cameras
The state of the art of deep learning in camera operations is reviewed and the impact of AI on the physical design of cameras is considered. Expand
DSCam: uma plataforma hardware-software para operações de visão computacional
O desenvolvimento de arquiteturas de processamento cada vez de menor consumo de energia e maior poder de processamento, aliado ao desenvolvimento de diversas outras tecnologias, possibilitou aExpand
Analog Devices , “ Blackfin processors