Analog VLSI and neural systems
Analog VLSI and neural systems
A floating-gate MOS implementation of resistive fuse
Neural Computation
A low-cost programmable clock generator for switched-capacitor circuit applications
Analog Integrated Circuits and Signal Processing
IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Hi-index | 0.00 |
In artificial vision systems edge detection combined with noise suppression is a commonly used first step. A known solution for this task is the resistive fuse algorithm. This paper presents a novel mixed signal implementation that combines the advantages of analog computing, i.e., low power consumption and small area usage, with scalability and limited programmability usually found in digital designs only. The presented prototype shows that massively parallel, analog image processing can be realized in a submicron, single-poly CMOS process without relying much on transistor properties. This makes design reuse of analog processing blocks feasible.