ISO 23150:2021 pdf download – Road vehicles — Data communication between sensors and data fusion unit for automated driving functions — Logical interface.
7 Object level 7.1 General The object level interfaces (OLI) originate from several sensors. Therefore, at object level, the term sensor cluster is always used, even if a single sensor is serving the interface. The OLIs provide recognised real-world entities and recognition is based on historical data. Recognised objects are tracked and filtered over time. Model-based algorithms classify these objects and rate them with an existence probability value. The sensor cluster refers its recognised objects to one of two predefined vehicle coordinate systems [see signal “Vehicle coordinate system type” (A.1.21)]. The properties, and therefore the signals of the recognised objects, are generally independent of the sensor cluster’s sensor technology. Additional properties of the recognised objects can depend on the sensor technology and complement the object description of the interface. These sensor technology dependent signals are grouped in LSGs for each sensor technology. To link recognised objects with non-OLI interface entities of the sensor cluster or its sensors, each recognised object has a unique signal’s “Object ID” (A.2.2) value that does not change over time and which is referenced by the entities. The signal’s “Object ID” (A.2.2) value can only be reused after the object has not been recognised for a defined amount of time and the sensor cluster is therefore certain that the object is no longer visible. Specifically, a reidentification of a previous recognised object with the same signal’s “Object ID” (A.2.2) value, after it has not been visible to the sensor cluster for a long time, is not guaranteed.
To link recognised objects of the sensor cluster’s OLIs, the set of related objects use the same consistent unique signal’s “Object grouping ID” (A.2.3) value to link all related objects. One dedicated value may indicate that the object is not linked to another object. The object recognition algorithm is based on multiple measurements of the sensor cluster and is updated over time. For this reason, the entire object properties are not assigned to the time of one single measurement, but to an estimated point in time at which the properties of the interface’s objects fit together consistently. The object existence probability relates to this consistent set of estimated signals of the object at this estimated point in time. On object level the following interfaces are available: — potentially moving object interface (7.3); — road object interface (7.4); — static object interface (7.5). 7.2 Generic object level interface Table 6 defines the generic structure of an OLI.