Disparity map building using guide node
An apparatus including an interface and a processor. The interface may be configured to receive pixel data. The processor may be configured to generate a reference image and a target image from the pixel data, perform disparity operations on the reference image and the target image and build a disparity map in response to the disparity operations. The disparity operations may comprise selecting a guide node from the pixel data comprising a pixel and a plurality of surrounding pixels, determining a peak location for the pixel by performing a full range search, calculating a shift offset peak location for each of the surrounding pixels by performing block matching operations in a local range near the peak location and generating values in a disparity map for the pixel data in response to the peak location for the pixel and the shift offset peak location for each of the surrounding pixels.
1. An apparatus comprising:
an interface configured to receive pixel data; and
a processor configured to (i) generate a reference image and a target image from said pixel data, (ii) perform disparity operations on said reference image and said target image and (iii) build a disparity map in response to said disparity operations, wherein said disparity operations comprise (a) selecting a guide node from said pixel data comprising a guide node pixel and a plurality of surrounding pixels, (b) determining a guide node pixel peak location for said guide node pixel by performing a full range search in said reference image and said target image, (c) calculating a shift offset peak location for each of said surrounding pixels in said guide node by performing block matching operations in a local range near said guide node pixel peak location and (d) generating values in said disparity map for said pixel data in said guide node in response to (i) said guide node pixel peak location for said guide node pixel and (ii) said shift offset peak location for each of said surrounding pixels.
2. The apparatus according to claim 1 , wherein said shift offset peak location for said surrounding pixels is determined in response to (i) selecting said local range near said guide node pixel peak location, (ii) generating coefficients in response to said block matching operations at three locations in said local range by comparing blocks of said pixel data near one of said surrounding pixels in said reference image and said target image, (iii) calculating said shift offset peak location in response to said coefficients and (iv) repeating (i)-(iii) for each of said surrounding pixels.
3. The apparatus according to claim 2 , wherein (i) said coefficients define a paracurve and (ii) said shift offset peak location is determined in response to calculating a peak location of said paracurve.
4. The apparatus according to claim 1 , wherein said full range search comprises (i) shifting locations of a block of pixels comprising said guide node pixel across a search range, (ii) calculating one of a plurality of matching scores for one of a plurality of locations of said block of pixels comprising said guide node pixel in said reference image and a corresponding location in said target image, (iii) repeating (i)-(ii) for each of said plurality of locations in said search range and (iv) selecting a highest value from said matching scores as said guide node pixel peak location.
5. The apparatus according to claim 4 , wherein said local range is smaller than said search range.
6. The apparatus according to claim 1 , wherein said disparity operations are configured to generate said disparity map in less time than using said block matching operations over a full video frame and with fewer mismatches than semi-global block matching operations.
7. The apparatus according to claim 1 , wherein said disparity operations are configured to build said disparity map at a sub-pixel level of accuracy.
8. The apparatus according to claim 1 , wherein said disparity operations comprise post-processing to prevent waveform distortions in said disparity map.
9. The apparatus according to claim 8 , wherein said post-processing comprises implementing a paracurve in order to calculate a matching location.
10. The apparatus according to claim 1 , wherein said disparity operations are configured to generate said disparity map at least 20 times faster than using said block matching operations over a full video frame.
11. A method for determining disparity values from stereo images, comprising the steps of:
receiving pixel data at an interface;
generating a reference image and a target image from said pixel data using a processor;
performing disparity operations using said processor on said reference image and said target image; and
building a disparity map in response to said disparity operations using said processor, wherein said disparity operations comprise (a) selecting a guide node from said pixel data comprising a guide node pixel and a plurality of surrounding pixels, (b) determining a guide node pixel peak location for said guide node pixel by performing a full range search in said reference image and said target image, (c) calculating a shift offset peak location for each of said surrounding pixels in said guide node by performing block matching operations in a local range near said guide node pixel peak location and (d) generating values in a said disparity map for said pixel data in said guide node in response to (i) said guide node pixel peak location for said guide node pixel and (ii) said shift offset peak location for each of said surrounding pixels.
12. The method according to claim 11 , wherein said shift offset peak location for said surrounding pixels is determined in response to said processor performing the steps of:
(i) selecting said local range near said guide node pixel peak location,
(ii) generating coefficients in response to said block matching operations at three locations in said local range by comparing blocks of said pixel data near one of said surrounding pixels in said reference image and said target image,
(iii) calculating said shift offset peak location in response to said coefficients, and
(iv) repeating (i)-(iii) for each of said surrounding pixels.
13. The method according to claim 12 , wherein (i) said coefficients define a paracurve and (ii) said shift offset peak location is determined in response to calculating a peak location of said paracurve.
14. The method according to claim 11 , wherein said full range search comprises said processor performing the steps of:
(i) shifting locations of a block of pixels comprising said guide node pixel across a search range,
(ii) calculating one of a plurality of matching scores for one of a plurality of locations of said block of pixels comprising said guide node pixel in said reference image and a corresponding location in said target image,
(iii) repeating (i)-(ii) for each of said plurality of locations in said search range, and
(iv) selecting a highest value from said matching scores as said guide node pixel peak location.
15. The method according to claim 14 , wherein said local range is smaller than said search range.
16. The method according to claim 11 , wherein said disparity operations are configured to generate said disparity map in less time than using said block matching operations over a full video frame and with fewer mismatches than semi-global block matching operations.
17. The method according to claim 11 , wherein said disparity operations are configured to build said disparity map at a sub-pixel level of accuracy.
18. The method according to claim 11 , wherein said disparity operations comprise post-processing to prevent waveform distortions in said disparity map.
19. The method according to claim 18 , wherein said post-processing comprises implementing a paracurve in order to calculate a matching location.
20. The method according to claim 11 , wherein said disparity operations are configured to generate said disparity map at least 20 times faster than using said block matching operations over a full video frame.