The Hopfield neural network as a tool for feature tracking and recognition from satellite sensor images

Abstract

A new approach for feature tracking and recognition on sequential satellite sensor images using neural networks has been developed. Feature tracking is recognized as being of importance in applications such as ice-mapping, cloud motion winds, ocean currents, and short-term forecasting. Feature recognition ® nds application in automatic image navigation. This paper explores the potential of a Hop® eld neural network to perform feature tracking or recognition, and gives examples of its implementation to three di erent applications. It is shown that the net can provide superior performance to existing techniques for tracking, the advantages of this new approach being its precision, speed, low sensitivity to deformation, its capacity to detect rotational motion, and to provide directly the cross-and along-isopycnal components of displacement vectors. It is also shown that the Hop® eld neural network can provide a valuable tool for automatic image navigation through coastline recognition.

Notice of Retraction: Application of optical flow method in inversion of ocean surface flow field

“…: detecting ocean flow by observing the motion of infrared picture temperature (Emery et al 1986;Kelly 1989;Cote and Tatnall 1995;Bowen et al 2002Crocker et al, 2007. Such method is called the Characteristic Quantity Tracing Method.…”

Cloud motion detection from infrared satellite images

“…The tree arches are assigned a cost based on the resemblance of considered blocks but also on their motion. Therefore, we have considered a weighted sum defined as Cost=pS+(l-p)C. (5) where the weight p is balancing the two factors significance.…”

Intelligent Global Vision for Teams of Mobile Robots

“…Neural networks have also been applied, though less extensively, to tracking information over time (e.g. (Cote and Tatnall, 1997)), and this also supports their use as a viable choice in real-time robot tracking.…”