US 12,254,603 B2
Adaptive wavelet denoising
Mohammad Izadi, San Jose, CA (US); Pavan Madhusudanarao, San Jose, CA (US); and Balineedu Adsumilli, Sunnyvale, CA (US)
Assigned to GOOGLE LLC, Mountain View, CA (US)
Appl. No. 17/908,068
Filed by GOOGLE LLC, Mountain View, CA (US)
PCT Filed May 19, 2020, PCT No. PCT/US2020/033655
§ 371(c)(1), (2) Date Aug. 30, 2022,
PCT Pub. No. WO2021/236070, PCT Pub. Date Nov. 25, 2021.
Prior Publication US 2023/0119747 A1, Apr. 20, 2023
Int. Cl. G06T 5/70 (2024.01); G06T 5/20 (2006.01)
CPC G06T 5/70 (2024.01) [G06T 5/20 (2013.01); G06T 2207/20021 (2013.01); G06T 2207/20064 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for processing image data, comprising:
generating, for an input image in a spatial domain, two-dimensional (2-D) wavelet coefficients at a plurality of levels, each level of the plurality of levels comprising multiple subbands each associated with a respective subband type in a wavelet domain;
for respective levels of the plurality of levels, identifying a flat region of the subband of the multiple subbands comprising blocks of the subband having a variance no higher than a first threshold variance;
identifying a flat block set for the subband type associated with the subband comprising blocks that are common to respective flat regions of the subband;
determining a second threshold variance using variances of the blocks of the flat block set;
thresholding, using the second threshold variance, at least some of the 2-D wavelet coefficients at the plurality of levels to remove noise; and
after thresholding, generating a denoised image in the spatial domain using the plurality of levels.