Search

This paper presents an algorithm for a team of heterogeneous mobile robots to estimate and adaptively sample a stationary, isotropic, Gaussian process. An estimation framework is proposed to assimilate measurements from robots with differing sensing capabilities (i.e., measurement noise variance). To improve computational efficiency of the Gaussian process regression, the survey area is divided into regions that each use a common semivariogram matrix constructed with a truncated measurement set determined by an adaptive selector. As the mission proceeds, a Voronoi-based algorithm periodically partitions a density function—representing time-varying sampling priority—to identify high-value sampling locations. The path for each robot is modeled as an mechanical system: a sequence of masses (waypoints) are interconnected by springs and dampers and pulled towards Voronoi cell centroids. At each path planning cycle the robots are iteratively simulated with heterogeneous dynamics (e.g., speed, turn radius) following their respective waypoint paths as stiffness/damping parameters are adjusted to satisfy mission time constraints. Numerical simulations show that the proposed approach reduces mapping error when compared to non-adaptive lawnmower coverage of a survey area. The algorithm is demonstrated experimentally using two cooperating autonomous surface vessels to map the bathymetry in a section of Lake Norman near Charlotte, NC.

This study describes a network and its application methods for efficient detection and recognition of the depth-of-field(DoF) region blurred in the image by focusing and defocusing the camera. This approach uses a cross-correlation filter based on RGB color channels to efficiently extract DoF regions in images and construct a dataset for training in the convolutional neural network. A data pair corresponding to the image-DoF weight map is set using the data. The training data are from a DoF weight map extracted based on an image and cross-correlation filter. The loss function is modeled using the result of applying Gaussian derivatives of the image to improve the convergence rate efficiently in the network training phase. The DoF weight map obtained as a test result and proposed in this paper reliably extracted the DoF region in the input image. In addition, this study experimentally demonstrates that the proposed method can be used in various applications, such as non-photorealistic rendering, viewpoint tracking, object detection and recognition, optical character recognition, and adaptive sampling, that employ the user regions of interest.

Sparse-viewcomputed tomography (CT) imaging is a promising technique for reducing radiation dose and accelerating data acquisition in medical imaging. However, the challenges of handling a reduced number of projection views persist for both iterative estimation and deep neural reconstruction methods. In this paper, to address these challenges, we present a singular value decomposition-based adaptive sampling approximate message passing network (ASAMP-Net) sparse-view CT imaging method. To achieve multiple sparse views projection within a single scene imaging and alleviate the computational burden, our proposed ASAMP-Net method incorporates an adaptive sampling module into the AMP deep unrolling network. This module dynamically adjusts the data samples used during the learning process, making our method highly adaptable to various projection matrices. Moreover, by decomposing the projection matrix into its principal components, our approach identifies the respective contributions of independent structures. We then select the most significant principal components to construct a projection matrix model with increased orthogonality, thereby enhancing reconstruction performance. Extensive experiments on public datasets demonstrate the superiority of our method. Notably, ASAMP-Net handles various sparse projection views with just a single training process, achieving prominent imaging results compared to other methods in the literature.