This sounds very promising! Lenseless cameras/imaging could be a game changer! However, this seems to be early stage research.
This computer vision research was not published in one of the usual journals.
"... The lensless camera’s conventional optical hardware comprises a thin mask and an image sensor. The mask and sensor can be produced combined in established semiconductor manufacturing techniques for future production. The mask optically encodes the incident light, which casts patterns on the sensor. ...
“Without the restrictions of a lens, the lensless camera might be ultra-miniature, allowing novel uses that are beyond our imagination.” ..."
“Without the restrictions of a lens, the lensless camera might be ultra-miniature, allowing novel uses that are beyond our imagination.” ..."
From the abstract:
"A mask-based lensless camera optically encodes the scene with a thin mask and reconstructs the image afterward. ... Conventional model-based reconstruction approaches, which leverage knowledge of the physical system, are susceptible to imperfect system modeling. Reconstruction with a pure data-driven deep neural network (DNN) avoids this limitation, thereby having potential to provide a better reconstruction quality. However, existing pure DNN reconstruction approaches for lensless imaging do not provide a better result than model-based approaches. We reveal that the multiplexing property in lensless optics makes global features essential in understanding the optically encoded pattern. ... With this analysis, for the first time to the best of our knowledge, a fully connected neural network with a transformer for image reconstruction is proposed. The proposed architecture is better in global feature reasoning, and hence enhances the reconstruction. The superiority of the proposed architecture is verified by comparing with the model-based and FCN-based approaches in an optical experiment."
Image reconstruction with transformer for mask-based lensless imaging (no public access)
No comments:
Post a Comment