International audienceWe present an algorithm based on MRF modelling for motion detection in image sequences and give a modified version for implementation on analog resistive network. Energy minimization is realized by a network relaxing to its state of minimal power dissipation. It takes a few nanoseconds and replaces advantageously time consuming stochastic or suboptimal deterministic relaxation algorithms. The elementary cell of the network is presented along with the environment needed to feed it with the required inputs. Two network architectures are proposed, derived from CCD camera principle. Software simulations of a 128\texttimes128 network demonstrate the good behaviour of the modified algorithm on real sequences. Electrical simu...
We have designed, built and tested a number of analog CMOS VLSI circuits for computing 1-D motion fr...
Abstract—Cellular Nonlinear Networks (CNN) establish a theoretical framework in which programmable f...
The detection of discontinuities in motion, intensity, color, and depth is a well-studied but diffic...
International audienceIn this paper we present two analog networks that implement an algorithm for m...
International audienceThe implementation of a visual motion detection algorithm on an analog network...
International audienceThe main concern in image processing is the computation cost. Markov random fi...
International audienceThis paper describes the real-time implementation of a simple and robust motio...
Abstract—In this paper, we propose an analog integrated circuit to compute the motion field of a tim...
A methodological approach to the definition of nonlinear circuits for real-time image processing is ...
Analog models provide a novel framework for understanding and developing algorithms for computer vis...
Abstract—This paper analyses a MOS-based resistive network suitable for massively parallel image pro...
The real time computation of motion from real images using a single chip with integrated sensors is ...
Two digital LSI implementation methods for nonlinear resistive networks are proposed; one is for pix...
The real time computation of motion from real images using a single chip with integrated sensors is ...
We propose herein a motion detection artificial vision model which uses analog electronic circuits. ...
We have designed, built and tested a number of analog CMOS VLSI circuits for computing 1-D motion fr...
Abstract—Cellular Nonlinear Networks (CNN) establish a theoretical framework in which programmable f...
The detection of discontinuities in motion, intensity, color, and depth is a well-studied but diffic...
International audienceIn this paper we present two analog networks that implement an algorithm for m...
International audienceThe implementation of a visual motion detection algorithm on an analog network...
International audienceThe main concern in image processing is the computation cost. Markov random fi...
International audienceThis paper describes the real-time implementation of a simple and robust motio...
Abstract—In this paper, we propose an analog integrated circuit to compute the motion field of a tim...
A methodological approach to the definition of nonlinear circuits for real-time image processing is ...
Analog models provide a novel framework for understanding and developing algorithms for computer vis...
Abstract—This paper analyses a MOS-based resistive network suitable for massively parallel image pro...
The real time computation of motion from real images using a single chip with integrated sensors is ...
Two digital LSI implementation methods for nonlinear resistive networks are proposed; one is for pix...
The real time computation of motion from real images using a single chip with integrated sensors is ...
We propose herein a motion detection artificial vision model which uses analog electronic circuits. ...
We have designed, built and tested a number of analog CMOS VLSI circuits for computing 1-D motion fr...
Abstract—Cellular Nonlinear Networks (CNN) establish a theoretical framework in which programmable f...
The detection of discontinuities in motion, intensity, color, and depth is a well-studied but diffic...