Adriano Basile

An object oriented segmentation on analog CNN chip

This paper introduces a real-time object oriented segmentation algorithm, designed and implemented on a new type of mixed analog/digital chip based on the cellular neural/nonlinear network (CNN) paradigm. The fully parallel architecture of the CNN processes all the pixels of an image at the same time, so the time spent for the image segmentation is independent of the number of objects in the image. This implementation of the segmentation algorithm is shown to well satisfy the real-time requirements both as a stand-alone processing procedure, and as a module inside the MPEG-4 video coding standard. Finally, the general purpose characteristics of the CNN universal chip allow to use the algorithm introduced as an efficient pre-processing procedure for many interesting image/video stand-alone applications.

Image processing for medical diagnosis using CNN

Abstract Medical diagnosis is one of the most important area in which image processing procedures are usefully applied. Image processing is an important phase in order to improve the accuracy both for diagnosis procedure and for surgical operation. One of these fields is tumor/cancer detection by using Microarray analysis. The research studies in the Cancer Genetics Branch are mainly involved in a range of experiments including the identification of inherited mutations predisposing family members to malignant melanoma, prostate and breast cancer. In bio-medical field the real-time processing is very important, but often image processing is a quite time-consuming phase. Therefore techniques able to speed up the elaboration play an important rule. From this point of view, in this work a novel approach to image processing has been developed. The new idea is to use the Cellular Neural Networks to investigate on diagnostic images, like: Magnetic Resonance Imaging, Computed Tomography, and fluorescent cDNA microarray images.

CNN wave based computation for robot navigation on ACE16K

The CNN wave based computation is an approach for real time robot navigation in a complex environment based on the idea of considering the environment in which the robot moves as an excitable medium. Obstacles represent the sources of autowave generation. The waveform propagating in the CNN medium provide to the robot all the information to achieve an adaptive motion avoiding the obstacles. In this paper we implement entirely this strategy on the ACE16K CNN-chip.

Watermarking on CNN-UM for image and video authentication

In this paper a new approach to fragile watermarking technique is introduced. This problem is particularly interesting in the field of modern multimedia applications, when image and video authentication are required. The approach exploits the cellular automata suitability to work as pseudorandom pattern generators and extends the related algorithms under the framework of the cellular non-linear networks (CNNs). The result is a novel way to perform watermarking generation in real time, using the presently available CNN-universal chip prototypes. In this paper, both the CNN algorithms for fragile watermarking as well as on-chip experimental results are reported, confirming the suitability of CNNs to successfully act as real-time watermarking generators. The availability of CNN-based visual microprocessors allows to have powerful algorithms to watermark in real time images or videos for efficient smart camera applications. Copyright

CNN wave based computation for robot navigation planning

In this work a methodology for real-time robot navigation in a complex, dynamically changing environment, based on wave computation and implemented by cellular neural networks (CNNs) is introduced. The keypoint of the approach is to consider the environment in which the robot moves as an excitable medium. Obstacles and targets represent the source of autowave generation. The wavefronts propagating in the CNN medium provide to the robot all the information to achieve an adaptive motion avoiding the obstacles and directed to the target. In particular the paradigm of reaction-diffusion (RD) equations are used to implement a CNN-based wave computation for navigation control. Experimental results validating the approach are shown.

Implementation of Turing patterns for bio-inspired motion control

In this paper a CNN controlling the reactive behavior of a roving robot by means of Turing patterns is introduced. The Turing pattern represents the fixed-action pattern of the robot, while the initial conditions of the CNN are given by the sensor status. The approach is still valid when the number of sensors is high, being able to perform data fusion in real-time through analog parallel processing. An experiment using a small roving robot is presented to validate the approach.

A BIO-INSPIRED VISUAL FEEDBACK LOCOMOTION CONTROL BASED ON CNN UNIVERSAL MACHINE

In this paper a new framework for bio-inspired robot locomotion control, entirely based on Cellular Neural Networks (CNNs), is introduced. In fact, CNNs are employed both for generating locomotion patterns in a hexapod robot, and for its trajectory control via visual feedback. In the paper the latter problem will be emphasized, being the CNN locomotion generation problem already treated in literature. Feedback signals are images captured by a camera and processed in real time by a CNN used as analog image processor. The actual framework makes use of a traditional PC where a tool for the synchronization of the visual CNN chip and the robot control has been designed. This paper describes the methodology as an important stage for the study, definition and optimization of the overall control methodology. The results obtained reveal true real time capabilities and the PC interface can be easily substituted with a processing board to be integrated with the visual CNN and with the locomotion-devoted CNN, in order to constitute a unified integrated system for real time visual motion control in complex structures.

Visual feedback by using a CNN chip prototype system

Robot locomotion control passes through a series of sensors that, according to information from the environment, allow the robot to adapt, in real time, its locomotion scheme or trajectory. When the goal of the robot is to reach a target in a non-structured environment the best approach is visual control realized by a fast image processing system. Fast parallel image processing of the CNN-UM cP4000 chip prototype permits one to obtain good performance, even in a real time control problem. The robot controlled by the implemented CNN visual feedback has a hexapod configuration and its locomotion system is also implemented by a multi-layer CNN structure. In this paper a CNN approach for both locomotion generation and visual control of the bio-inspired robot is presented.

CNN-based real-time video detection of plasma instability in nuclear fusion applications

In this paper a real-time detection of plasma instabilities, called MARFEs, is performed through a real-time image processing on plasma video sequences. These sequences are recorded by a vision system based on a CCD camera installed at Frascati Tokamak Upgrade (FTU). The strategy used to perform the task is based on a new family of nonlinear analog processors, digitally programmable, implemented into the so-called cellular neural network universal machine (CNN-UM). The detection system allows to carry out safer nuclear fusion experiments, preventing the plant from excessive mechanical and thermal stress which occurs during plasma instability phenomena (i.e. disruptions). Experimental results, obtained on the FTU machine, are fully satisfactory.

Watermarking for the authentication of video on CNN-UM

Digital watermarks have been proposed for authentication of both video and still images. In such applications, the watermark is embedded within a host image such that subsequent alteration to the watermarked image can be detected with high probability. In this paper the possibility of implementing real time watermarking on a video stream is presented. In fact the new CNN-UM implementation offers time operation of only microseconds working on 64/spl times/64 images.