Search

Ubiquitous low cost multi-rotor and fixed wing drones or unmanned aerial vehicles (UAVs) have accelerated the need for reliable, robust, and scalable Structure-from-Motion (SfM) and Multi-View Stereo (MVS) pipelines suitable for a variety of flightpath trajectories especially in degraded environments. Feature tracking being a core part of SfM and MVS, is essential for multiview scene modeling and perception, but difficult to evaluate in large scale datasets due to the lack of sufficient ground-truth. For large-scale aerial imagery, accurate camera orientation and dense 3D point cloud accuracy can be used to assess the impact of accurate feature localization and track length. We propose a novel view simulation (or synthesis) framework which generates visually realistic new unseen camera views for feature detection using known high fidelity camera poses for modeling. Seven state-of-the-art local handcrafted and learning-based features are quantitatively evaluated for robustness and matchability within the SfM and MVS pipelines using the open source COLMAP software. Our experimental results provide performance rankings of each feature, using twelve different evaluation metrics across three synthetic city-wide aerial image sequences. We show that recent learned features, SuperPoint and LF-Net, have not only reached the quality of the best handcrafted features like SIFT and SURF, but now outperform them in terms of more accurate 3D camera pose estimates and longer feature tracks. SuperPoint produces 1.51 meter average position error and 0.03° average angular error, while SIFT remains competitive (second best for pose and overall) with 1.78 meter and 0.11° errors respectively. [ABSTRACT FROM AUTHOR]

We describe a fast and efficient camera pose refinement and Structure from Motion (SfM) method for sequential aerial imagery with applications to georegistration and 3-D reconstruction. Inputs to the system are 2-D images combined with initial noisy camera metadata measurements, available from on-board sensors (e.g., camera, global positioning system, and inertial measurement unit). Georegistration is required to stabilize the ground-plane motion to separate camera-induced motion from object motion to support vehicle tracking in aerial imagery. In the proposed approach, we recover accurate camera pose and (sparse) 3-D structure using bundle adjustment for sequential imagery (BA4S) and then stabilize the video from the moving platform by analytically solving for the image-plane-to-ground-plane homography transformation. Using this approach, we avoid relying upon image-to-image registration, which requires estimating feature correspondences (i.e., matching) followed by warping between images (in a 2-D space) that is an error prone process for complex scenes with parallax, appearance, and illumination changes. Both our SfM (BA4S) and our analytical ground-plane georegistration method avoid the use of iterative consensus combinatorial methods like RANdom SAmple Consensus which is a core part of many published approaches. BA4S is very efficient for long sequential imagery and is more than 130 times faster than VisualSfM, 35 times faster than MavMap, and about 274 times faster than Pix4D. Various experimental results demonstrate the efficiency and robustness of the proposed pipeline for the refinement of camera parameters in sequential aerial imagery and georegistration. [ABSTRACT FROM PUBLISHER]

We present a novel formulation of the problem of energy misbehavior and develop an analytical framework for quantifying its impact on other nodes. Specifically, we formulate two versions of the power control problem for wireless sensor networks with latency constraints arising from duty cycle allocations. In the first version, strategic power optimization, nodes are modeled as rational agents in a power game, who strategically adjust their powers to minimize their own energy. In the other version, joint power optimization, sensor nodes adjust their transmission powers to minimize the aggregate energy expenditure. Our analysis of these models yields insight into the different energy outcomes of strategic versus joint power optimization. We show that while joint power optimization fits the accepted paradigm of cooperation among sensor nodes (for example large number of sensor nodes cooperating for a task such as target tracking), it comes with both advantages and disadvantages when energy misbehavior is taken into account. One advantage is that it can (sometimes) be energy-dominant, i.e., the optimal energy cost for each node under joint energy minimization is lower than its strategically optimal energy cost. We then develop a model for characterizing energy misbehavior and show that joint optimization is disadvantageous because it is impossible to prevent misbehavior under any channel quality and load constraints, whereas strategic optimization is more resilient. We prove that it is impossible for a node to unilaterally and undetectably follow a different energy optimization strategy than the other nodes and hence the only threat to the network is misbehavior through false advertisement. We then provide sufficient conditions under which misbehavior through false advertisement can be prevented under a strategic optimization regime. Our analytical results reveal optimal strategies for attacking nodes in an enemy network through energy depletion and help develop effective defense mechanisms for protecting our own wireless network against energy attacks by an intelligent adversary. [ABSTRACT FROM AUTHOR]

We present a rigorous geometric analysis of the computation of the global positions of an airborne video camera and ground based objects using aerial images of known landmarks. This has also been known as the perspective-n-point (PnP) problem. A robust Hough transform-like method, facilitated by a class of CORDIC-structured computations is developed to find the camera position followed by a method of computing the position of a ground object from images of that object and three known landmarks. The results enable fast and effective visual terrain navigation of aerial surveillance systems when the global positioning and inertial navigation sensors become faulty, inaccurate, or dysfunctional. These new hardware implementable algorithms can also be used with MEMS based INS sensors through a multisensory fusion process. [ABSTRACT FROM AUTHOR]