Title :
Implementation of cellular neural network operating with bipolar images
Author :
Paasio, Ari ; Dawidziuk, Adam ; Porra, Veikko
Author_Institution :
Electron. Circuit Design Lab., Helsinki Univ. of Technol., Espoo, Finland
Abstract :
The paper presents a 16×16 cellular neural network chip implementation. The circuit has been processed with 1.2 micron technology. Cell nonlinearity is based on a high gain sigmoid and this allows processing of black and white images. The layout of the design is discussed and finally measurement results which show almost fully correct behavior are presented
Keywords :
CMOS analogue integrated circuits; analogue processing circuits; cellular neural nets; edge detection; integrated circuit layout; neural chips; piecewise-linear techniques; 1.2 micron technology; 1.2 mum; 16×16 cellular neural network chip implementation; analog CMOS; bipolar images; cell nonlinearity; high gain sigmoid; inverse edge detection; piecewise linear sigmoid; Cellular neural networks; Circuit topology; Electronic circuits; Electronic mail; Integrated circuit measurements; Laboratories; Paper technology; Semiconductor device measurement; Switches; Turing machines;
Conference_Titel :
Neural Networks, 1996., IEEE International Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-3210-5
DOI :
10.1109/ICNN.1996.549016
