Summary Maps for Lifelong Visual Localization

Abstract

Robots that use vision for localization need to handle environments that are subject to seasonal and structural change, and operate under changing lighting and weather conditions. We present a framework for lifelong localization and mapping designed to provide robust and metrically accurate online localization in these kinds of changing environments. Our system iterates between offline map building, map summary, and online localization. The offline mapping fuses data from multiple visually varied datasets, thus dealing with changing environments by incorporating new information. Before passing these data to the online localization system, the map is summarized, selecting only the landmarks that are deemed useful for localization. This Summary Map enables online localization that is accurate and robust to the variation of visual information in natural environments while still being computationally efficient. We present a number of summary policies for selecting useful features for localization from the multisession map, and we explore the tradeoff between localization performance and computational complexity. The system is evaluated on 77 recordings, with a total length of 30 kilometers, collected outdoors over 16 months. These datasets cover all seasons, various times of day, and changing weather such as sunshine, rain, fog, and snow. We show that it is possible to build consistent maps that span data collected over an entire year, and cover day‐to‐night transitions. Simple statistics computed on landmark observations are enough to produce a Summary Map that enables robust and accurate localization over a wide range of seasonal, lighting, and weather conditions.

Simultaneous Localization and Mapping: A Survey of Current Trends in Autonomous Driving

“…In [166], the authors consider that, in the future, maps will be coming from centralizing collect services. As such, they propose a system where maps, constructed using vision by optimizing over 2D-3D associations, serve to feed a database.…”

“…However, the information density needed for most approaches makes a world-wide deployment difficult to envisage at the moment. Direct availability of maps is the main problem but it is worth noting that [166] mentions probe cars (multiple sensorequipped vehicles) as a future way to reduce this problem. This will be further discussed in Section V.…”

“…While these approaches are only applied to a single-vehicle, the extension to a fleet is straightforward. In [166], the authors mention that the objective in the long-term is the deployment of such a system in probe cars that would gather data and fuse them in the cloud to then provide maps as a service. The main difficulties are the necessity to divide maps by sectors in order to be still optimizable in a reasonable time and to limit how maps can grow to avoid having intractable environment representation.…”

Multi-modal and inertial sensor solutions for navigation-type factor graphs

“…The required Entity-Relationship model for designing such databases assumes a specific mechanism of mapping and localization acting on the raw data, a mechanism which might not satisfy the interface needs of a new technique [161]. Moreover, these database methods require validity checking for each new datum and are not able to integrate new data into the map online [162].…”

“…Moreover, these database methods require validity checking for each new datum and are not able to integrate new data into the map online [162]. This method doesn't scale because of the necessary re-incorporation of all data for each query of metric computation, thus precluding online augmentation of the map [161].…”