All-optical neural-net-like image processing with photosensitive nonlinear nematic film

The equivalence of all-optical image processing with the optical nonlinearities of nematic liquid crystal to neural-net-like processing operations was demonstrated. The photosensitivity and transparency of the film allowed integration of the sensor, processor and image display function on a single film. An example of image edge enhancement was illustrated with photosensitive supranonlinear nematic liquid-crystal films. The results show that some neural-net image-processing operations can be realized by use of the natural diffraction property of light in conjunction with the optical nonlinearities of the transparent film.