In this paper, we present an analytical design approach for the class of bipolar cellular neural networks (CNN's) which yields optimally robust template parameters. We give a rigorous definition of absolute and relative robustness and show that all well-defined CNN tasks are characterized by a finite set of linear and homogeneous inequalities. This system of inequalities can be analytically solved for the most robust template by simple matrix algebra. For the relative robustness of a task, a theoretical upper bound exists and is easily derived, whereas the absolute robustness can be arbitrarily increased by template scaling. A series of examples demonstrates the simplicity and broad applicability of the proposed method.

: In this paper, a CNN based spatio-temporal approach is introduced for finding the endocardial (inner) boundary of the left ventricle from a sequence of echocardiographic images. The discussed analogic 1 CNN algorithm combines optimal nonlinear filtering and constrained wave propagation in order to estimate the continuous contour of a moving object in a medium where the edges are ill-defined. In the preprocessing phase, nonlinear filtering is employed to remove the coherent speckle noise that corrupts the images. It is verified that an optimal filtering strategy should estimate the mode of the local intensity histogram. Three different approximations of the mode filter were implemented, derived from robust statistics 2 and geometry-driven diffusion 3 , that give an output consistent with the maximum likelihood estimate of the noisy sequence. The kernel of the left ventricle is located and the boundary is found using a fuzzy-adaptive technique that embodies constrained w...

: This paper is concerned with the implementation of cryptographic hash functions on the regular array of simple cellular neural network (CNN, [1, 2]) cells with periodic boundary conditions. Cryptographic hash functions enable message origin authentication and validation of message content integrity. A class of cryptographic hash functions - termed Cartesian authentication codes - provide provable (unconditional) security for message authentication between two mutually trustful parties sharing a secret key. We succeeded in implementing existing constructions of Cartesian authentication codes on today's CNN Universal Machine (CNN-UM) chips [3, 4]. Here we prove that rather complex (binary) arithmetic can be performed on a simple CNN chip, by providing an algorithm to implement a specific Cartesian authentication code based on the computation of a polynomial expression over a finite field. The bitrate of the computation is in the 100Mbit/sec range with existing chips. 1. Introduction C...

Cellular Neural Networks are very suitable to realize single-chip focal-plane image processing for artificial vision in the context of space-variant active vision. A new solution is presented for optical flow computation in continuous time, by a differential method based on the equation of continuity and the smoothness constraint. I. INTRODUCTION A. Space-Variant Active Vision and CNNs Artificial perception is drawing growing interest, with projected application in numerous fields, in particular for robotics and, in the long term, for the design of prosthetic devices. Strict constraints of size, weight, cost and performance must be jointly optimized in these applications, and for this reason, based on inspiration from biology, the concept of space-variant active vision (SVAV) [1] has been developed in the last decade and is now recognized as a very promising approach. SVAV is based on the use of sensors characterized by a smooth variation of resolution across the workspace, like that...

: In this paper a spatio-temporal analogic CNN algorithm is designed for front-end filtering, segmentation and object recognition. First, a generalized segmentation strategy is presented based on various diffusion models. Both PDE and non-PDE related schemes are discussed and their VLSI complexity is analyzed. In classification (object recognition) a CNN implementation of the autowave metric, a "nonlinear" variant of the Hausdorff metric, is used. This approach turned out to be superior compared to some other classification methods, e.g. the Hamming distance calculation. A number of tests have been completed within the so-called "bubble/debris" segmentation experiments using original and artificial gray-scale images. 1. Introduction Since the publication of the original paper in 1988 ([1]), the rapidly growing field of Cellular Neural Networks (CNNs) have found numerous potential applications, especially in image processing problems where real-time signal processing is required. The ...

Abstract not found

In this thesis a new signal processing technique is presented. This technique exploits the use of pulses as the signalling mechanism. This Palmo 1 signalling method applied to signal processing is novel, combining the advantages of both digital and analogue techniques. Pulsed signals are robust, inherently low-power, easily regenerated, and easily distributed across and between chips. The Palmo cells used to perform analogue operations on the pulsed signals are compact, fast, simple and programmable.