Signal Analysis and Classification for Surface Plasmon Assisted Microscopy of Nanoobjects

The need for portable and on site screening methods for viruses is evident in face of virus infections that can spread fastly in a heavily connected world. A robust and efficient method for detecting viruses is a novel technique called Plasmon Assisted Microscopy of Nanoobjects. It is based on the acquisition of images from a sensor surface exploiting the behavior of surface plasmons in the presence of nanoobjects. In this paper an efficient image analysis approach with respect to the requirements of the sensor is presented and an embedded image processing system for this purpose is introduced. The processing pipeline comprises three steps and starts with restorating the images by removing the background and filtering artifacts. The acquired image series is analyzed pixel by pixel in a second pipeline step in order to detect pixels containing nanoobjects. In a last step pixels are aggregated to nanoobject structures. The paper introduces in the context of this virus detection method a configurable embedded system that was used for rapid prototyping of the image analysis algorithms in a flexible way.