Methods circuits devices systems and functionally associated machine executable code for generating a searchable real-scene database
This application discloses to methods, circuits, devices, assemblies and systems for generating a searchable real-scene database including records indicating entities and physical events occurring within the real-scene. The said system may include a camera feed interface to receive a camera feed from each of one or more cameras observing the real-scene. It may also include image processing circuitry including a static scene analysis unit to: (a) extract features of entities appearing in a camera feed, (b) recognize entity types of entities appearing in the one or more camera feeds, (c) assign an entity designator to a specific entity appearing in the camera feeds, (d) generate an entity designator record for a uniquely identifiable entity, and generate an attribute record for a uniquely identifiable entity. The system may also include a scene dynamics analysis unit to: (a) detect entity movements, (b) detect entity speech, (c) characterize entity actions based on detected entity movement and/or entity speech, and (d) generate a record indicting an action taken by an entity.
1. A system for generating a searchable real-scene database including records indicating entities and physical events occurring within the real-scene, said system comprising:
a camera feed interface to receive a camera feed from each of one or more cameras observing the real-scene;
an image processing circuitry including:
(1) a static scene analysis unit to:
a. extract features of entities appearing in a camera feed;
b. recognize entity types of entities appearing in the one or more camera feeds;
c. assign an entity designator to a specific entity appearing in the camera feeds;
d. generate an entity designator record for a uniquely identifiable entity;
e. generate an attribute record for a uniquely identifiable entity;
(2) a scene dynamics analysis unit to:
a. detect entity movements;
b. detect entity speech;
c. characterize entity actions based on detected entity movement or entity speech; and
d. generate a record indicting an action taken by an entity;
a scene data query engine to search and retrieve records from said digital data storage in accordance with a user query received through a public user interface;
a real-time scene query interface to receive a user scene query to be resolved using real-time camera feeds; and
a real-time scene query controller to in invoke and control one or more scene camera(s) while assessing at least one output from said scene analysis units in accordance with a user scene query,
wherein said controller includes camera control logic to perform a scan of the real-scene with the one or more camera(s) in order to resolve the user scene query and
wherein said camera control logic is adapted to use scene related data records stored in said digital data storage in order to focus the scan of the real-scene when attempting to resolve the user scene query.
2. The system according to claim 1 , wherein said scene dynamics analysis unit is further adapted to detect and record a relationship two or more between entities based on detected interactions between the two or more entities.
3. The system according to claim 1 , wherein an identifier is generated and linked to an entity designator of a specific entity based on natural language processing of speech spoken by another entity addressing the specific entity.
4. The system according to claim 1 , further comprising scene data controller including: (a) cross domain digital data storage devices adapted to store data of various types and data records indicating various physical events and entities appearing in the stored data; and (b) a cross domain digital data storage controller adapted to index and log data records into said digital data storage devices.
5. The system according to claim 4 , wherein said scene data controller further comprises a scene data query engine to search and retrieve records from said digital data storage in accordance with a user query received through a system user interface.
6. The system according to claim 1 , further comprising a scene data query engine to search and retrieve records from said digital data storage in accordance with a user query received through a private system user interface.
7. The system according to claim 1 , wherein said public user interface is functionally associated with an abstractor to abstract or otherwise generifying user queries received through said public user interface.
8. The system according to claim 1 , wherein said public user interface is functionally associated with anonymizer to anonymize or otherwise generifying system responses to user queries received through said public user interface.
9. The system according to claim 1 , wherein failure to resolve a received user scene query using scene data records stored in the digital data storage triggers use of the real-time scene query interface.
10. The system according to claim 1 , wherein said real-time scene query controller is further adapted to invoke one or more search assisting robot(s) functionally associated with the system to move around and acquire video information from the real-scene according to patterns intended to resolve the user scene query.
11. The system according to claim 1 , further comprising an audio processor adapted to characterize sound and speech generated by entities within the real-scene.
12. The system according to claim 11 , wherein said audio processor is adapted to perform natural language processing on speech generated by a first entity interacting with a second entity.
13. The system according to claim 12 , wherein said audio processor is synchronized with an entity action detector.
14. The system according to claim 13 , wherein said audio processor is adapted to detect a name used by the first entity when referring to the second entity.
15. The system according to claim 14 , wherein said audio processor provides the detected name to generate an attribute record for the second entity.