The Technical use of Mobile Robots in Developing Smart Cities in Integration with Vins

The handling of independent mobile robots in complex settings presents a difficult task. Overcoming problems related to incorrect labeling and frequent tracking loses, particularly in difficult scenes, goes beyond the powers of point-based visual SLAM and simultaneous mapping (vSLAM) methods. This study presents a creative real-time, alongside resilient point-line oriented monocular visual inertial Navigation (VINS) system designed for mobile robots in urban areas that are ready for 6 G advances. To enhance resilience in tracking difficult scenes, we implement EDLines alongside adaptive kappa correction for swift capture of a higher share of extended line attributes within all derived line features. A new real-time line feature-mapping technique is presented to track this retrieved line features between subsequent frames minus the necessity about computing descriptors, giving an increase of three in performance compared with the traditional LSD and KNN matching strategy based on LBD descriptors. Moreover, a tightly linked sensor fusion optimization scheme is built to attain precise state estimates, covering point-line characteristic reprojection errors and accelerometer device (IMU) residuals. Through tests on freely available standard datasets, our VINS system shows better localization precision real-time performance, and stability when compared to other modern SLAM systems. This VINS system enables the accurate location of robots on the move in smart cities defined by complicated surroundings.