Cooperative remote sensing of ice using a Spatially Indexed Labeled Multi-Bernoulli filter

Abstract

Abstract-In polar region operations, drift ice positioning and tracking is useful for both scientific and safety reasons. Many sensors can be employed to generate detections of sea ice, such as satellite-carried Synthetic Aperture Radar (SAR) and, recently, imagery equipment carried by Unmanned Aerial Systems (UAS). Satellite-carried SAR has the advantage of being able to cover large areas and provide consistent imagery largely independent of weather, albeit at a relatively coarse resolution. Using UAS, the resolution and precision of the tracking can be locally improved.To track the large amount of individual objects present in an area as large as the Arctic, it is necessary to efficiently select and exclusively work with the objects in the relevant field-of-view. In this paper, a Spatially Indexed Labeled Multi-Bernoulli filter is presented and applied to a tracking problem representing a mission setup for field-tests due this year. In the setup, satellite and UAS imagery is combined to provide real-time Multi-Target Tracking of sea ice objects. A brief introduction is given to the implementation of the proposed Spatially Indexed Labeled Multi-Bernoulli filter, which is made available under an Open Source license.

Outline of a system for integrated adaptive ice tracking and multi-agent path planning

“…Partial results of the different subsystems have been published previously in [17][18][19], but are recapitulated here as a descriptive example of the accomplished work.…”

“…In [17,19], we introduced the spatially indexed Labeled Multi-Bernoulli (lmb) filter for scalable tracking of sea ice. The lmb implementation was used to track the ice movements in Kongsfjorden, Svalbard, over a period of seven hours, with scans delivered every three minutes.…”

“…Sea ice coverage mapping can be performed using occupancy grids [14] whereas identifiable individual objects are tracked with sparse target tracking [7,16,17,19]. In the former case, scalability can be achieved through variable resolution, whereas in the latter, objects can be stored in efficient datastructures [16,17] to limit the number of objects involved in each update. An example result of the data fusion, from [19], is found in Figure 3.…”