US 12,254,635 B2
Image distortion correction method and apparatus
Chen Su, Shenzhen (CN); Huanwen Peng, Shenzhen (CN); Shuiran Peng, Shanghai (CN); Yihua Zeng, Shenzhen (CN); and Shisheng Zheng, Hangzhou (CN)
Assigned to HUAWEI TECHNOLOGIES CO., LTD., Shenzhen (CN)
Filed by Huawei Technologies Co., Ltd., Shenzhen (CN)
Filed on Mar. 18, 2022, as Appl. No. 17/698,650.
Application 17/698,650 is a continuation of application No. PCT/CN2020/114400, filed on Sep. 10, 2020.
Claims priority of application No. 201910896549.6 (CN), filed on Sep. 18, 2019.
Prior Publication US 2022/0207747 A1, Jun. 30, 2022
Int. Cl. G06T 7/12 (2017.01); G06T 3/18 (2024.01); G06T 3/40 (2024.01); G06T 3/60 (2024.01); G06T 7/194 (2017.01)
CPC G06T 7/12 (2017.01) [G06T 3/18 (2024.01); G06T 3/40 (2013.01); G06T 3/60 (2013.01); G06T 7/194 (2017.01); G06T 2207/20132 (2013.01); G06T 2207/30196 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An image distortion correction method, comprising:
performing optical distortion correction on a collected source image to obtain a first corrected image, wherein the first corrected image comprises a background region, a field of view edge region, and a portrait with stretching deformation, and wherein the portrait comprises at least a first human body region and a second human body region;
performing algorithm constraint correction on the first human body region, the second human body region, and the background region by using constraint terms to obtain a second corrected image, wherein the constraint terms for performing the algorithm constraint correction respectively constrain the first human body region, the second human body region, and the background region;
constructing the constraint terms, wherein the constraint terms further constrain the field of view edge region;
constructing a regular constraint term for the first corrected image;
obtaining displacement matrices respectively corresponding to the first human body region, the second human body region, and the background region based on the constraint terms and the regular constraint term;
obtaining target image matrices respectively corresponding to the first human body region, the second human body region, and the background region based on initial image matrices and the displacement matrices; and
obtaining the second corrected image through color mapping based on the target image matrices and color matrices respectively corresponding to the first human body region, the second human body region, and the background region.