Fig 1.
The branching structures of trees generated by a L-system.
(a) The structure with four iterations. (b) The structure with six iterations.
Fig 2.
The 3D model of a maple tree visualized in a CAD viewer.
(a) The branch structures, with vertices of surface meshes plotted as a point cloud. (b) The tree leaves plotted as triangular meshes.
Fig 3.
Samples from IPPs with a mixture of squared exponential intensity function (p = 2).
(a) A sample corresponds to C1 = 1.0, C2 = 0.6, (a1, b1) = (5, 5), (a2, b2) = (15, 15), and (h1, l1) = (h2, l2) = (3, 2). (b) A sample corresponds to C1 = 0.4, C2 = 0.2, (a1, b1) = (5, 5), (a2, b2) = (10, 10), (h1, l1) = (3, 4), and .
Fig 4.
A comparison between simulated and measured echoes for a maple tree.
(a) An echo signal measured from a maple tree. (b) An simulated echo signal obtained by matching the simulation setup with the experimental setup. (c) Plots of the empirical CDFs of the signal amplitudes.
Fig 5.
The simulation results of a single tree and a forest.
(a) The first level branches generated by the L-system. (b) Sub-branches are added. (c) A view of the complete tree with leaves plotted as green dots. (d) The 3D visualization of a forest with two species of trees.
Fig 6.
A static sonar with varying beamwidths at 10, 30, 50 degrees (left to right). Red regions highlight the -3 dB contour of the sonar beam. Waveforms of the impulse responses were plotted on the left side of the 2-D view below the sonar location.
Fig 7.
A UAV navigates across a simulated forest following a “8” shaped route. Twelve sampling points are taken along the route. Sonar beamwidth varies. The -3 dB contour of the sonar beams are highlighted in red. The waveforms of impulse responses are plotted at each sampling location.
Table 1.
The computation time under two setups.