METHOD AND SYSTEM FOR FAST DENSE STEREOSCOPIC RANGING
First Claim
1. A stochastic method for fast stereoscopic ranging comprising:
- (a) selecting a pair of images for stereo processing, wherein the pair of images are a frame pair and one of the image is a reference frame;
(b) seeding estimated values for a range metric at each pixel of the reference frame;
(c) initializing one or more search stage constraints;
(d) stochastically computing local influence for each valid pixel in the reference frame, wherein a valid pixel is a pixel in the reference frame that has a corresponding pixel in the other frame of the frame pair;
(e) aggregating local influences for each valid pixel in the reference frame;
(f) refining the estimated values for the range metric at each valid pixel in the reference frame based on the aggregated local influence; and
(g) post-processing range metric data; and
(h) repeating n iterations of (d) through (g), wherein n is a value selected in (c).
1 Assignment
0 Petitions
Accused Products
Abstract
A stochastic method and system for fast stereoscopic ranging includes selecting a pair of images for stereo processing, in which the pair of images are a frame pair and one of the image is a reference frame, seeding estimated values for a range metric at each pixel of the reference frame, initializing one or more search stage constraints, stochastically computing local influence for each valid pixel in the reference frame, aggregating local influences for each valid pixel in the reference frame, refining the estimated values for the range metric at each valid pixel in the reference frame based on the aggregated local influence, and post-processing range metric data. A valid pixel is a pixel in the reference frame that has a corresponding pixel in the other frame of the frame pair. The method repeats n iterations of the stochastically computing through the post-processing.
-
Citations
23 Claims
-
1. A stochastic method for fast stereoscopic ranging comprising:
-
(a) selecting a pair of images for stereo processing, wherein the pair of images are a frame pair and one of the image is a reference frame; (b) seeding estimated values for a range metric at each pixel of the reference frame; (c) initializing one or more search stage constraints; (d) stochastically computing local influence for each valid pixel in the reference frame, wherein a valid pixel is a pixel in the reference frame that has a corresponding pixel in the other frame of the frame pair; (e) aggregating local influences for each valid pixel in the reference frame; (f) refining the estimated values for the range metric at each valid pixel in the reference frame based on the aggregated local influence; and (g) post-processing range metric data; and (h) repeating n iterations of (d) through (g), wherein n is a value selected in (c). - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A method for initializing depth estimates in an iterative process for stereoscopic ranging, the method comprising:
-
projecting a first depth estimate for each one of a plurality of valid pixels in an image, the image having an implicit camera origin; generating a 3D representation of the plurality of valid pixels using the first depth estimate; relocating the implicit camera origin; reprojecting the 3D representation of the plurality of valid pixels to the relocated implicit camera origin; using Z-buffering to render a second depth estimate for each one of the plurality of valid pixels according to the relocated implicit camera origin; and using the second depth estimate for each of the plurality of valid pixels as an initial value in a subsequent iteration of a process for stereoscopic ranging. - View Dependent Claims (19)
-
-
20. A computer readable medium comprising instructions for performing a stochastic method for fast stereoscopic ranging by:
-
(a) selecting a pair of images for stereo processing, wherein the pair of images are a frame pair and one of the image is a reference frame; (b) seeding estimated values for a range metric at each pixel of the reference frame; (c) initializing one or more search stage constraints; (d) stochastically computing local influence for each valid pixel in the reference frame, wherein a valid pixel is a pixel in the reference frame that has a corresponding pixel in the other frame of the frame pair; (e) aggregating local influences for each valid pixel in the reference frame; (f) refining the estimated values for the range metric at each valid pixel in the reference frame based on the aggregated local influence; (g) post-processing range metric data; and (h) repeating (d) through (g) for n iterations, wherein n is a value selected in (c).
-
-
21. A system for stochastic fast stereoscopic ranging comprising:
-
a visual sensor capable of taking images; a computing device coupled to the visual sensor, comprising a memory, wherein the memory includes instructions for performing a stochastic method for fast stereoscopic ranging by; (a) selecting a pair of images for stereo processing, wherein the pair of images are a frame pair and one of the image is a reference frame; (b) seeding estimated values for a range metric at each pixel of the reference frame; (c) initializing one or more search stage constraints; (d) stochastically computing local influence for each valid pixel in the reference frame, wherein a valid pixel is a pixel in the reference frame that has a corresponding pixel in the other frame of the frame pair; (e) aggregating local influences for each valid pixel in the reference frame; (f) refining the estimated values for the range metric at each valid pixel in the reference frame based on the aggregated local influence; (g) post-processing range metric data; and (h) repeating (d) through (g) for n iterations, wherein n is a value selected in (c). - View Dependent Claims (22, 23)
-
Specification