US 12,260,575 B2
Scale-aware monocular localization and mapping
Benjamin Busam, Munich (DE); and Gregory Slabaugh, Munich (DE)
Assigned to HUAWEI TECHNOLOGIES CO., LTD., Guangdong (CN)
Filed by HUAWEI TECHNOLOGIES CO., LTD., Guangdong (CN)
Filed on Oct. 28, 2021, as Appl. No. 17/513,596.
Application 17/513,596 is a continuation of application No. PCT/EP2019/061071, filed on Apr. 30, 2019.
Prior Publication US 2022/0051425 A1, Feb. 17, 2022
Int. Cl. G06K 9/00 (2022.01); G06T 3/18 (2024.01); G06T 3/40 (2006.01); G06T 7/55 (2017.01); G06T 7/60 (2017.01); G06T 7/73 (2017.01)
CPC G06T 7/55 (2017.01) [G06T 3/18 (2024.01); G06T 3/40 (2013.01); G06T 7/60 (2013.01); G06T 7/73 (2017.01); G06T 2207/10016 (2013.01); G06T 2207/10021 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An image processing device, comprising:
a processor configured to perform operations of estimating a scale of image features, the operations comprising:
processing a plurality of images of a scene using a first trained model to identify features in the plurality of images and to estimate depths of the features in the plurality of images, wherein the plurality of images comprise a first set of images captured at a first location and a second set of images captured at a second location;
processing the plurality of images using a second trained model to estimate a scaling parameter for the images based on a feature difference in feature space between a first feature map corresponding to the first set of images and a second feature map corresponding to the second set of images; and
estimating scales of the features by adjusting the estimated depths based on the estimated scaling parameter.