Multi-sensor depth mapping

Depth mapping apparatus, comprising:

• an illumination assembly, which is configured to direct modulated optical radiation toward a target scene;
• a camera, which is configured to capture a two-dimensional image of the target scene;
• a range sensor, which is configured to sense respective times of flight of photons reflected from a matrix of locations disposed across the target scene;
• and
• a processor, which is configured to derive first depth coordinates of the matrix of locations responsively to the respective times of flight, to derive second depth coordinates of the matrix of locations responsively to a transverse disparity between features in the two-dimensional image and corresponding reference features in a reference image, to compute a disparity correction function based on a difference between the first and the second depth coordinates at the matrix of locations, to correct the transverse disparity between the two-dimensional image and the reference image using the disparity correction function, and to generate a depth map of the target scene based on the corrected transverse disparity,
• wherein the processor is configured to compute a range correction function over a first span of distances from the apparatus by comparing the first depth coordinates to the second depth coordinates in the depth map, and to apply the range correction function in correcting the first depth coordinates over a second span of distances, greater than the first span.

A method for depth mapping, comprising:

• directing modulated optical radiation toward a target scene;
• capturing a two-dimensional image of the target scene;
• sensing respective times of flight of photons reflected from a matrix of locations disposed across the target scene;
• deriving first depth coordinates of the matrix of locations responsively to the respective times of flight;
• deriving second depth coordinates of the matrix of locations responsively to a transverse disparity between features in the two-dimensional image and corresponding reference features in a reference image;
• computing a disparity correction function based on a difference between the first and the second depth coordinates at the matrix of locations,
• wherein computing the disparity correction function comprises calculating respective local disparity corrections at the matrix of the locations and interpolating and extrapolating from the local disparity corrections to generate the disparity correction function over an entire field of view of the two-dimensional image;
• correcting the transverse disparity between the two-dimensional image and the reference image using the disparity correction function;
• and
• generating a depth map of the target scene based on the corrected transverse disparity.