US 12,094,093 B2
Image descattering method based on iterative optimization of atmospheric transmission matrix
Xin Jin, Guangdong (CN); and Yihui Fan, Guangdong (CN)
Assigned to Tsinghua Shenzhen International Graduate School, Guangdong (CN)
Filed by Tsinghua Shenzhen International Graduate School, Guangdong (CN)
Filed on Mar. 18, 2022, as Appl. No. 17/697,948.
Application 17/697,948 is a continuation of application No. PCT/CN2021/123321, filed on Oct. 12, 2021.
Claims priority of application No. 202011337360.2 (CN), filed on Nov. 25, 2020.
Prior Publication US 2022/0207661 A1, Jun. 30, 2022
Int. Cl. G06T 5/00 (2024.01); G06T 5/50 (2006.01); G06T 5/73 (2024.01); G06T 7/521 (2017.01)
CPC G06T 5/73 (2024.01) [G06T 5/50 (2013.01); G06T 7/521 (2017.01); G06T 2207/10028 (2013.01); G06T 2207/20132 (2013.01); G06T 2207/20221 (2013.01); G06T 2207/30192 (2013.01)] 10 Claims
OG exemplary drawing
 
1. An image descattering method based on iterative optimization of an atmospheric transmission matrix, comprising the following steps:
S1, constructing a descattering model based on the atmospheric transmission matrix;
S2, estimating a forward scattering coefficient q corresponding to a foggy day image B;
S3, based on a depth map T of the foggy day image B and the forward scattering coefficient q obtained in the step S2, estimating an initial atmospheric transmission matrix A; and
S4, substituting the estimated initial atmospheric transmission matrix A into the descattering model in the step S1, and performing iterative update under a constraint condition, until a value of the descattering model satisfies a convergence condition, to obtain an optimal atmospheric transmission matrix A* and an optimal descattered image X.