.png){kind=icon}
.png){kind=icon}
Return
A novel dynamic tracking method for coded targets with complex background noise
.png){kind=icon}
PRE
AI
DOI：10.1016/j.optlaseng.2024.108654.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
To address the issues of low tracking efficiency and poor localization accuracy of artificial coded targets under complex background interference conditions, a new method for dynamic tracking of coded targets is proposed. This method includes a lightweight feature tracker (CBAM-Slim-Net) for adaptive localization of coded circles and a large-capacity coded target solver (CSN-BSSCT). The CBAM-Slim-Net feature tracker achieves a detection accuracy of 0.987 with only 6.030 M parameters. In actual measurement environments with complex background interference, CSN-BSSCT can decode quickly and accurately, with a mean error of 0.036 mm in the threedimensional Euclidean distance between coded circles in static measurement scenarios. Additionally, this method can analyze the motion trajectory of the target and perform dynamic stitching for 3D measurement from multiple perspectives, making it highly significant for applications in robot motion control and large-field-of-view 3D measurement.
Keywords:
CSN-BSSCT
Background noise
Dynamic tracking
CBAM-Slim-Net
Three-dimensional measurement
Journal
IF:
3.7
Papers:
7.1K
Citations:
1.7W
Organization
View more organizations