An error diagnosis device diagnosing if location measurement error has occurred in a location measurement sensor measuring a self-location of a vehicle has a storing part storing map information divided in every road sections; a location acquiring part acquiring self-location information of the vehicle measured by the location measurement sensor; a drive section identifying part identifying road sections on which the vehicle has been driven in the map information, based on the self-location information; and an error diagnosis part. The error diagnosis part judges that there is location measurement error in the location measurement sensor when a first and second road sections of one of the road sections identified as having been driven on by the vehicle are not consecutive, and judges that there is no location measurement error in the location measurement sensor when the first road section and the second road section are consecutive.

Apparatuses, systems, and techniques to perform trajectory predictions within one or more images. In at least one embodiment, a processor comprises one or more circuits to cause one or more neural networks to perform trajectory predictions of two or more objects detected within a plurality of frames without tracking the two or more objects based, at least in part, on processing a sequence of data of the one or more objects as a whole.

In one embodiment, a method includes at a mapping server receiving from a plurality of regional servers a plurality of regional magnetic data sets; at the mapping server updating a geomagnetic map with one or more of the regional magnetic data sets, wherein the geomagnetic map comprises one or more of the geographic regions; and from the mapping server communicating one or more portions of the geomagnetic map as updated to one or more of the regional servers for distribution as geomagnetic map information to one or more of the magnetic navigation devices for localization or navigation.

A method for detecting changes in road information includes converting, by a domain adapter, a captured road image and a projected road-layout map into first intermediate data and second intermediate data of a same feature space, respectively, calculating, by a similarity determiner, a similarity between the captured road image and the projected road-layout map based on the first intermediate data and the second intermediate data, and detecting, by a change detector, presence or absence of changes in road information on the projected road-layout map based on the calculated similarity.

An apparatus for managing maps is provided. The apparatus includes a processor configured to estimate a usage frequency of each of maps stored in a memory included in a vehicle and used for automated driving control of the vehicle, based on section information indicating sections used for automated driving control in each of the maps; and delete, from the memory, a map whose usage frequency is less than a frequency threshold.

Disclosed are a preprocessing method and device for distance transformation. The method includes: acquiring a first grid map; calculating a second parameter of a vehicle model; calculating a precision of a second grid map according to the second parameter of the vehicle model and a precision of the first grid map; calculating the numbers of rows and columns of the second grid map according to information on the first grid map, the precision of the second grid map and the second parameter of the vehicle model; determining for each cell of the second grid map a state value according to the numbers of obstacle cells within said each cell; and determining in the first grid map at least one cell requiring no processing, according to the state value. Therefore, during the distance transformation, the processing speed for determining the distance value of each cell is increased.

In one embodiment, a method includes at a mapping server receiving from a plurality of regional servers a plurality of regional magnetic data sets; at the mapping server updating a geomagnetic map with one or more of the regional magnetic data sets, wherein the geomagnetic map comprises one or more of the geographic regions; and from the mapping server communicating one or more portions of the geomagnetic map as updated to one or more of the regional servers for distribution as geomagnetic map information to one or more of the magnetic navigation devices for localization or navigation.

The present disclosure provides methods, apparatuses, and storage media for storing and loading visual localization maps. The methods for storing the visual localization maps may include acquiring a visual localization map; extracting key frame abstract information of each key frame from the visual localization map; grouping the key frame abstract information of all key frames of the visual localization map; for each group, generating and storing a sub-map-file of the group using the key frame abstract information of the group; generating key frame space index information based on the key frame abstract information of all groups; and generating and storing a master map file according to the key frame space index information for indexing the sub-map-file. By adopting the above solution, it is easy to flexibly load and manage the visual localization map according to the application requirements, thereby greatly improving the scalability of the visual localization map, the loading efficiency in the application and the space efficiency during the execution.

Disclosed are a preprocessing method and device for distance transformation. The method includes: acquiring a first grid map; calculating a second parameter of a vehicle model; calculating a precision of a second grid map according to the second parameter of the vehicle model and a precision of the first grid map; calculating the numbers of rows and columns of the second grid map according to information on the first grid map, the precision of the second grid map and the second parameter of the vehicle model; determining for each cell of the second grid map a state value according to the numbers of obstacle cells within said each cell; and determining in the first grid map at least one cell requiring no processing, according to the state value. Therefore, during the distance transformation, the processing speed for determining the distance value of each cell is increased.

A high-definition map building method includes determining a key node describing information about a key position of a lane attribute change, determining a key node layer based on a position of the key node and an attribute of the key node, and determining a high-definition map based on a navigation map and the key node layer. The navigation map provides road-level navigation information, and the high-definition map provides lane-level navigation information.