High-precision and high-robustness visual inertial odometer positioning method based on multi-camera observation
The invention provides a high-precision and high-robustness visual inertial odometer positioning method based on multi-camera observation, which belongs to the technical field of computer vision, and comprises the following steps of: firstly, carrying out synchronous data acquisition of multiple cam...
Saved in:
Main Authors | , , , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
25.06.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention provides a high-precision and high-robustness visual inertial odometer positioning method based on multi-camera observation, which belongs to the technical field of computer vision, and comprises the following steps of: firstly, carrying out synchronous data acquisition of multiple cameras, carrying out multi-camera front-end feature tracking on each image in an image group, and obtaining a feature association result of each image group; based on this, a cost function is constructed, by constructing a three-layer adaptive confidence weighting model, adaptive multi-camera back-end state estimation is realized, and sliding window optimization is executed. According to the method, the three-layer adaptive confidence weighting model capable of adapting to scene change is constructed, so that the dynamic adjustment of the camera weight is realized, the multi-camera data is fused to the maximum extent and effectively utilized, the feature quality, the tracking error and the reliability of the camera o |
---|---|
Bibliography: | Application Number: CN202410350840 |