Object-independent piston diagnosing approach for segmented optical mirrors via deep convolutional neural network

Abstract

Piston diagnosing approaches based on neural networks have shown great success while a few methods areheavily dependent on the imaging target of the optical system. In addition, they are inevitably faced with theinterference of submirrors. Therefore, a unique object-independent feature image is used to form an originalkind of data set. Besides, an extremely deep image-based convolutional neural network (CNN) of 18 layers isconstructed. Furthermore, 9600 images are generated as data set for each submirror with special measure ofsensitive area extracting. The diversity of results among all the submirrors is also analyzed to ensuregeneralization ability. Finally, the average root mean square error of six submirrors between the real pistonvalues and the predicted values is approximately 0.0622λ. Our approach has the following characteristics: (1)the data sets are object-independent and contain more effective details, which behave comparatively better inCNN training; (2) the complex network is deep enough and only a limited number of images are required; (3)the method can be applied to the piston diagnosing of segmented mirror to overcome the difficulty broughtby the interference of submirrors. Our method does not require special hardware, and is fast to be used atany time, which may be widely applied in piston diagnosing of segmented mirrors.