These are the attributes for public objects in aerial drone images. These are:

Common Metadata Attributes for Cataloged Objects

• Geographic coordinates (latitude, longitude, altitude)

• Timestamp of detection (date, time)

• Object class/type (e.g., tree, car, lamp post, mailbox, bridge, ship, etc.)

• Detection confidence score (probability or model output for accuracy)

• Bounding box geometry

o Point or centroid coordinates

o Bounding box dimensions (width, height, possibly orientation for “oriented bounding boxes” [OBB])

o Polygon outline for irregular or large structures

• Size/scale attributes (e.g., trunk diameter for trees, vehicle size, pool area)

• Appearance features

o Color/texture descriptors

o Feature vector/embedding from CNN or deep learning model

• Source image/reference

o ID of the aerial image and corresponding street view or map images

o Camera parameters (view angle, altitude, lens metadata)

• Semantic map context

o Proximity to roads, buildings, other map layers

• Spatial context/neighbor relationships

o Neighboring detected objects (for spatial analysis, clustering, duplicate removal)

• Viewpoint and multi-view metadata

o The set of images (street/aerial) where the object is visible

• Environmental conditions

o Weather, lighting, season—for monitoring temporal changes

• Object status or change flags (e.g., newly appeared, moved, changed, removed)

• Unique object ID

o Persistent identifier assigned during cataloging or deduplication

• Auxiliary attributes

o Object-specific parameters: For trees, may include species, health status; for vehicles, type/brand (if inferable)

• Privacy/Classifiers

o Flags for sensitive/private or public object (where regulations apply)

These attributes go far beyond just timestamp and location, powering advanced filtering, sorting, and user queries for analytics and applications. Cataloging systems as described in knowledge graph-based frameworks also record relationships (edges) between objects and contextual labels for richer semantic search and retrieval.