16 March 2017

Device Zenytsa

Zenytsa  is an optical device that provides compact and cheap way  to make distant objects appear magnified on images without huge complicated long-focus lenses. The possible difference in weight in cost is more than 100 times for 300mm lens.

Commonly a long focus objective consists of about 10 lenses with total weight between 2 and 3 kg.  We propose only one, easy to use simple lens that weighs just 0.02 kg, has 10 mkm thickness and innovative software method for sharp image reconstruction.

Apart from unique configuration of the fresnel lens we provide the special method of  distortion correction based on robust convolutional neural network. Our methods give exceptional results in obtaining extra quality images in reasonable time. The nature of the distortions is nonlinear. However, the best nonlinear filter for image reconstruction is our deep convolutional neural network, which gives us the better reconstruction without any visible artifacts. Fringing artifact is present in the window of the middle picture (usual filtering) and it is absent in the left picture, obtained by deep convolutional network.

Our product is ready to revolutionize the entire computer vision industry because of  its compact size and low price. It would be especially helpful for a robotic vision, cctv cameras, a 3D reconstruction for “smart city” projects  and many other areas of expertise.

30 years of experience in the field of Image Processing Systems Institute of RAS, Moscow, together with Samara University experts in the field of diffractive optics with the guidance of academician Soifer, Dr. Kazansky  and talented engineers of Zenytsa team made this product become possible.

Dr. Kazansky Nikolay Lvovich

Academician Soifer Victor Aleksandrovich