No abstract

The three-dimensional (3D) perception of autonomous vehicles is crucial for localization and analysis of the driving environment, while it involves massive computing resources for deep learning, which can't be provided by vehiclemounted devices. This requires the use of seamless, reliable, and efficient massive connections provided by the 6G network for computing in the cloud. In this paper, we propose a novel deep learning framework with 6G enabled transport system for joint optimization of depth and ego-motion estimation, which is an important task in 3D perception for autonomous driving. A novel loss based on feature map and quadtree is proposed, which uses feature value loss with quadtree coding instead of photometric loss to merge the feature information at the texture-less region. Besides, we also propose a novel multi-level V-shaped residual network to estimate the depths of the image, which combines the advantages of V-shaped network and residual network, and solves the problem of poor feature extraction results that may be caused by the simple fusion of low-level and high-level features. Lastly, to alleviate the influence of image noise on pose estimation, we propose a number of parallel sub-networks that use RGB image and its feature map as the input of the network. Experimental results show that our method significantly improves the quality of the depth map and the localization accuracy and achieves the state-of-the-art performance.

No abstract

The expressions of wave structure function (WSF) and long-exposure modulation transfer function (MTF) for laser beam propagation through non-Kolmogorov turbulence were derived in our previous work. In this paper, based on anisotropic maritime atmospheric non-Kolmogorov spectrum, the new analytic expression of WSF for Gaussian-beam waves propagation through turbulent atmosphere in a horizontal path is derived. Moreover, using this newly derived expression, long-exposure MTF for Gaussian-beam waves is obtained for analyzing the degrading effects in an imaging system. Using the new expressions, WSF and MTF for Gaussian-beam waves propagating in terrestrial and maritime atmospheric turbulence are evaluated. The simulation results show that Gaussian-beam waves propagation through maritime turbulence obtain more degrading effects than terrestrial turbulence due to the humidity and temperature fluctuations. Additionally, the degrading effects under anisotropic turbulence get less loss than that of isotropic turbulence.

For the free-space optical (FSO) communication system, the spatial coherence of a laser beam is influenced obviously as it propagates through the atmosphere. This loss of spatial coherence limits the degree to which the laser beam is collimated or focused, resulting in a significant decrease in the power level of optical communication and radar systems. In this work, the analytic expressions of wave structure function for plane and spherical wave propagation through anisotropic non-Kolmogorov turbulence in a horizontal path are derived. Moreover, the new expressions for spatial coherence radius are obtained considering different scales of atmospheric turbulence. Using the newly obtained expressions for the spatial coherent radius, the effects of the inner scales and the outer scales of the turbulence, the power law exponent, and the anisotropic factor are analyzed. The analytical simulation results show that the wave structure functions are greatly influenced by the power law exponent α , the anisotropic factor ζ , the turbulence strength σ ~ R 2 , and the turbulence scales. Moreover, the spatial coherence radiuses are also significantly affected by the anisotropic factor ζ and the turbulence strength σ ~ R 2 , while they are gently influenced by the power law exponent α and the inner scales of the optical waves.