No abstract

Recently, the philosophy of visual saliency and attention has started to gain popularity in the robotics community. Therefore, this paper aims to mimic this mechanism in SLAM framework by using saliency prediction model. Comparing with traditional SLAM that treated all feature points as equal important in optimization process, we think that the salient feature points should play more important role in optimization process. Therefore, we proposed a saliency model to predict the saliency map, which can capture both scene semantic and geometric information. Then, we proposed Salient Bundle Adjustment by using the value of saliency map as the weight of the feature points in traditional Bundle Adjustment approach. Exhaustive experiments conducted with the state-of-the-art algorithm in KITTI and EuRoc datasets show that our proposed algorithm outperforms existing algorithms in both indoor and outdoor environments. Finally, we will make our saliency dataset and relevant source code open-source for enabling future research.

No abstract

For manually driven rubber trams to track, virtual tracks can easily cause driver fatigue. Therefore, based on visual navigation, an automatic steering and trajectory following method are proposed. First, the vehicle kinematic and dynamic model of the Delight Tram is proposed. Then, the automatic steering and trajectory following methods are introduced, which are based on model prediction control and Ackermann steering theory, respectively. Finally, the effectiveness of the proposed methods has been evaluated via both multi-body dynamic simulations and road tests under various working conditions. The results show that the vehicle has excellent steering and trajectory following ability whether in a transient phase or a steady-state circumference. Furthermore, the steering system can stabilize the vehicle in the whole range of design speed, with a smaller computational cost.

Objective To explore navigation-related factors interfering with accuracy of robot-assisted surgery. Methods We made a measurement model to test the accuracy of the TianJi Robot system when performing the stimulated screw placement procedure. The three-coordinate machine was used to measure the deviation between the actual position and the planned position. We designed corresponding experiments to explore the effects of different navigation-related factors on the screw placement accuracy. The deviations were measured at different distance (ranging from 1.2 m to 2.2 m) between the navigation optical stereo camera and the tracker and each distance was measured 50 times. The distance between the optical camera and the patient tracker was set at 1.4 m and the deviations were measured at different angles between the camera and the robot tracker, each angle was measured more than 25 times. Data was donated with mean and standard deviation. The line charts were employed to describe the changes of deviations over one clinical factor including distance and angle. Results Within the available scope of navigation optical system (1.2 m-2.2 m), the deviation increased with the distance (χ2=479.107, P<0.001). The robotic system accuracy was high and stable (mean deviation 0.332 mm ± 0.067 mm) when the relative angle between the optical camera and the tracker less than 40 degrees. Conclusions Accuracy of robot system was affected by the relative distance and angle between the optical camera and the tracker. When placing and adjusting the optical tracking devices, surgeons should set the relative distance between the optical camera and the patient tracker as 1.4 m- 1.5 m and the relative angle less than 40 degrees.