Hybrid depth imaging with sparse subject irradiation
A method enacted in a depth-imaging system comprises (a) modulating radiant output from an emitter at one or more modulation frequencies; (b) projecting the radiant output as a plurality of sparse-projection features; (c) acquiring a plurality of raw shutters on an imaging sensor array modulated at the one or more modulation frequencies; (d) triangulating a geometric estimate of depth to a subject locus reflecting a sparse-projection feature; and (e) returning a time-of-flight value of the depth based on a phase computed from the plurality of raw shutters and on the geometric estimate of the depth.
1 . A method enacted in a depth imaging system, the method comprising:
modulating radiant output from an emitter at one or more modulation frequencies, the emitter arranged optically upstream of a redistribution optic configured to sparsely project the radiant output onto a subject;
modulating charge-carrier collection at an imaging sensor array at the one or more modulation frequencies;
acquiring a plurality of raw shutters of the subject on the imaging sensor array;
constructing a phase map of the subject based on the plurality of raw shutters;
pinpointing in the phase map a plurality of bright areas corresponding each to a sparse-projection feature of the radiant output reflecting from the subject; and
for each of the plurality of bright areas:
triangulating an estimate of depth to the corresponding sparse-projection feature reflecting from the subject, wherein the estimate is based on a displacement of the bright area from a calibration position along a trajectory,
computing an aggregate phasor, and
phase unwrapping the aggregate phasor based partly on the estimate, to reveal a depth value.
2 . The method of claim 1 wherein each sparse-projection feature of the radiant output comprises a dot.
3 . The method of claim 1 wherein the aggregate phasor is associated with a set of periodic depth values, and wherein the phase unwrapping selects the depth value from the set.
4 . The method of claim 1 wherein pinpointing the plurality of bright areas comprises pinpointing based on zeroeth- and first-order image moments.
5 . The method of claim 1 wherein pinpointing the plurality of bright areas comprises pinpointing via multi-scale Laplacian-of-Gaussian detection.
6 . The method of claim 1 wherein pinpointing the plurality of bright areas comprises computing a centroid of each of the plurality of bright areas.
7 . The method of claim 1 wherein the estimate is inversely proportional to the displacement.
8 . The method of claim 7 further comprising acquiring and storing a calibration image, wherein a constant of proportionality relating the estimate to the displacement is based in part on the calibration image.
9 . The method of claim 1 wherein measuring the displacement includes matching each bright area to a corresponding trajectory.
10 . The method of claim 9 wherein the trajectories corresponding to adjacent bright areas do not overlap.
11 . The method of claim 10 wherein the aggregate complex intensity for each non-saturating bright area is computed via a joint bilateral filter.
12 . The method of claim 9 wherein the trajectories corresponding to adjacent bright areas overlap, and wherein the matching is based in part on the aggregate phasor.
13 . A depth-imaging system comprising:
an emitter configured to emit a modulated radiant output;
a redistribution optic arranged optically downstream of the emitter and configured to sparsely project the radiant output onto a subject;
an imaging sensor array configured to acquire a plurality of raw shutters of the subject;
a computer configured to:
modulate the radiant output of the emitter and a charge-carrier collection at the imaging sensor at one or more modulation frequencies,
construct a phase map of the subject based on the plurality of raw shutters,
pinpoint in the phase map a plurality of bright areas corresponding each to a sparse-projection feature of the radiant output reflecting from the subject; and
for each of the plurality of bright areas: triangulate an estimate of depth to the corresponding sparse-projection feature reflecting from the subject, wherein the estimate is based on a displacement of the bright area from a calibration position along a trajectory, compute an aggregate phasor, and phase unwrap the aggregate phasor based partly on the estimate, to reveal a depth value.
14 . The depth-imaging system of claim 13 wherein each sparse-projection feature of the radiant output comprises a dot.
15 . The depth-imaging system of claim 13 wherein the aggregate phasor is associated with a set of periodic depth values, and wherein the phase unwrapping selects the depth value from the set.
16 . The depth-imaging system of claim 13 wherein the estimate is based on a separation between adjacent bright areas.
17 . A method enacted in a depth-imaging system, the method comprising:
modulating radiant output from an emitter at one or more modulation frequencies;
projecting the radiant output as a plurality of sparse-projection features;
acquiring a plurality of raw shutters on an imaging sensor array modulated at the one or more modulation frequencies;
triangulating a geometric estimate of depth to a subject locus reflecting a sparse-projection feature, wherein the geometric estimate of depth is based on a displacement of a bright area in a phase map derived from the plurality of raw shutters from a calibration position along a trajectory, the bright area corresponding to the sparse-projection feature reflected from the subject locus; and
returning a time-of-flight value of the depth based on a phase computed from the plurality of raw shutters and on the geometric estimate of the depth.
18 . The method of claim 17 wherein each sparse-projection feature of the radiant output comprises a dot.
19 . The method of claim 17 wherein the one or more modulation frequencies comprise only two modulation frequencies.
20 . The method of claim 17 wherein the one or more modulation frequencies comprise only one modulation frequency.