A needle-based deep-neural-network camera
This enables imaging in confined spaces, with potential applications in medical endoscopy and privacy-enhanced classification, though it is incremental as it applies existing DNN methods to a new optical setup.
The authors tackled the problem of imaging through a narrow cannula by using deep neural networks to reconstruct color, grayscale, and depth maps from transported light intensity, achieving a field of view of 180 and angular resolution of ~0.40.
We experimentally demonstrate a camera whose primary optic is a cannula (diameter=0.22mm and length=12.5mm) that acts a lightpipe transporting light intensity from an object plane (35cm away) to its opposite end. Deep neural networks (DNNs) are used to reconstruct color and grayscale images with field of view of 180 and angular resolution of ~0.40. When trained on images with depth information, the DNN can create depth maps. Finally, we show DNN-based classification of the EMNIST dataset without and with image reconstructions. The former could be useful for imaging with enhanced privacy.