Ultrafast light field tomography

A light field tomography imaging method, comprising:

• (a) providing a light field tomography apparatus comprising:
• (i) a cylindrical lens with an invariant axis, the lens configured to produce parallel beam projections of an object in an image plane parallel to the invariant axis of the cylindrical lens;
• (ii) an array of cylindrical lenslets, each lenslet with an invariant axis oriented at a distinct angle;
• and
• (iii) an array of Dove prisms positioned between the cylindrical lens and the array of lenslets, each Dove prism optically coupled with a lenslet of the array of cylindrical lenslets;
• (iv) an imaging device positioned after the array of lenslets, said device configured to image said parallel beam projections from each lenslet;
• (b) acquiring parallel beam projections of a target by the imaging device;
• (c) ordering the acquired beam projections having spatial axis and distinct angles (x, θ) into sinograms;
• and
• (d) reconstructing 2D (x, y) images from the sinograms.

A light field tomography imaging method, comprising:

• (a) providing a light field tomography apparatus comprising:
• (i) an imaging device having an entrance slit and a lens;
• and
• (ii) a plurality of cylindrical lenslets positioned in a cylindrical lenslet array wherein each lenslet is oriented at a distinct angle;
• (iii) the cylindrical lenslet array positioned between the lens and entrance slit and aligned with the entrance slit, wherein the lenslets are arranged in a sequence of their invariant axis' angles with respect to the entrance slit;
• (b) imaging a 3D target by the imaging device lens to an intermediate image space, from which the cylindrical lenslet array forms differently projected sub-images onto a plane of the entrance slit;
• and
• (c) using a field stop in the intermediate image plane to reduce a field of view to avoid potential sub-image overlap between adjacent lenslets;
• (d) wherein the imaging device relays 1D projection images extracted by the entrance slit onto a photocathode, converting images to an electronic domain, and deflects the images onto different rows of a charge-coupled device (CCD) detector according to time of arrival of photons.

A light field tomography (LIFT) apparatus, comprising:

• (a) a cylindrical lens with an invariant axis, the lens configured to produce parallel beam projections of an object in an image plane parallel to the invariant axis of the cylindrical lens;
• (b) an array of cylindrical lenslets, which is separate from the cylindrical lens, wherein each lenslet of the array of cylindrical lenslets has an invariant axis oriented at a distinct angle, and positioned in line with the cylindrical lens;
• and
• (c) an imaging device positioned after the array of lenslets, with the array of cylindrical lenslets positioned between the cylindrical lens and the imaging device, wherein said imaging device is configured to image said parallel beam projections which pass through each lenslet of the array of cylindrical lenslets into said imaging device;
• (d) wherein said lenslets of the array of cylindrical lenslets are arranged in a sequence of their invariant axis' angles with respect to the imaging device;
• (e) an array of Dove prisms, each Dove prism optically coupled with a lenslet of the array of cylindrical lenslets;
• (f) a target illumination light source, wherein reflected light from a target is imaged by the imaging device;
• and
• (g) wherein a single snapshot image of the object is produced by the imaging device from simultaneously acquired beam projections from the sequence of lenslets.

A light field tomography (LIFT) apparatus, comprising:

• (a) a cylindrical lens with an invariant axis, the lens configured to produce parallel beam projections of an object in an image plane parallel to the invariant axis of the cylindrical lens;
• (b) an array of cylindrical lenslets, which is separate from the cylindrical lens, wherein each lenslet of the array of cylindrical lenslets has an invariant axis oriented at a distinct angle, and positioned in line with the cylindrical lens;
• and
• (c) an imaging device positioned after the array of lenslets, with the array of cylindrical lenslets positioned between the cylindrical lens and the imaging device, wherein said imaging device is configured to image said parallel beam projections which pass through each lenslet of the array of cylindrical lenslets into said imaging device;
• (d) wherein said lenslets of the array of cylindrical lenslets are arranged in a sequence of their invariant axis' angles with respect to the imaging device;
• (e) a processor configured to control said lenses and imaging device;
• and
• (f) a non-transitory memory storing instructions executable by the processor;
• (g) wherein said instructions, when executed by the processor, perform steps comprising:
• (i) acquiring a dataset of image data from the imaging device over time;
• (ii) reconstructing 2D, 3D or 4D images from the acquired dataset;
• (iii) applying a trained deep adjoint neural network (DANN) to the reconstruction;
• and
• (iv) displaying the reconstructed image on a display.