Real-time system for multi-modal 3D geospatial mapping, object recognition, scene annotation and analytics
A multi-sensor, multi-modal data collection, analysis, recognition, and visualization platform can be embodied in a navigation capable vehicle. The platform provides an automated tool that can integrate multi-modal sensor data including two-dimensional image data, three-dimensional image data, and motion, location, or orientation data, and create a visual representation of the integrated sensor data, in a live operational environment. An illustrative platform architecture incorporates modular domain-specific business analytics “plug ins” to provide real-time annotation of the visual representation with domain-specific markups.
1. A mobile computing device, comprising:
one or more processors;
one or more image sensors in communication with the one or more processors, the one or more image sensors configured to obtain multi-dimensional image data including at least one of two-dimensional image data and three-dimensional image data; and
one or more non-transitory machine accessible storage media in communication with the one or more processors, the one or more non-transitory machine accessible storage media comprising instructions to cause the mobile computing device to perform recognition of a plurality of visual features based on a map representation of a geo-spatial area of real world surroundings of the mobile computing device generated based on temporal and spatial alignment of the multi-dimensional image data, and
wherein the recognition of the plurality of visual features includes recognition of larger-scale objects, recognition of smaller-scale objects by performing context-free object identification and contextual object identification, and recognition of a complex object comprising a plurality of the smaller-scale objects.
2. The mobile computing device of claim 1 , wherein, the one or more non-transitory machine accessible storage media comprise instructions to cause the mobile computing device to temporally and spatially align the multi-dimensional image data and generate the map representation of the geo-spatial area of the real world surroundings of the mobile computing device based on the temporally and spatially aligned multi-dimensional image data.
3. The mobile computing device of claim 1 , wherein the recognition of the complex object is performed using a classifier.
4. The mobile computing device of claim 1 , wherein the recognition of the larger-scale objects is performed by determining a contextual frame of reference, and using the contextual frame of reference to identify the larger-scale objects.
5. The mobile computing device of claim 1 , wherein the recognition of the larger-scale objects is performed by executing an invariant multi-dimensional feature detection algorithm on data from at least one of the image sensors.
6. The mobile computing device of claim 1 , wherein the recognition of the smaller-scale objects is performed by executing a context-free feature-sharing algorithm.
7. The mobile computing device of claim 6 , wherein the recognition of the smaller-scale objects is performed by obtaining context information and classifying the smaller-scale objects based on the context information.
8. The mobile computing device of claim 7 , wherein the recognition of the complex objects is performed by executing a contextual bag of objects algorithm.
9. An object/scene recognition system comprising instructions embodied in one or more non-transitory computer readable storage media executable by one or more processors to cause a mobile computing device to:
perform recognition of a plurality of visual features based on a map representation of a geo-spatial area of real world surroundings of the mobile computing device generated based on temporal and spatial alignment of multi-dimensional image data obtained by one or more sensors,
wherein the multi-dimensional image data includes at least one of two-dimensional image data and three-dimensional image data, and
wherein the recognition of the plurality of visual features includes recognition of larger-scale objects, recognition of smaller-scale objects by performing context-free object identification and contextual object identification, and recognition of a complex object comprising a plurality of the smaller-scale objects.
10. The object/scene recognition system of claim 9 further comprising instructions embodied in the one or more non-transitory computer readable storage media executable by one or more processors to obtain the multi-dimensional image data from the one or more image sensors, perform temporal and spatial alignment of the multi-dimensional image data, and generate the map representation of the geo-spatial area of the real world surroundings of the mobile computing device based on the temporally and spatially aligned multi-dimensional image data.
11. The object/scene recognition system of claim 9 , wherein the recognition of the complex object is performed using a classifier.
12. The object/scene recognition system of claim 9 , wherein the recognition of the larger-scale objects is performed by determining a contextual frame of reference, and using the contextual frame of reference to identify the larger-scale objects.
13. The object/scene recognition system of claim 9 , wherein the recognition of the larger-scale objects is performed by executing an invariant multi-dimensional feature detection algorithm on data from at least one of the image sensors.
14. The object/scene recognition system of claim 9 , wherein the recognition of the smaller-scale objects is performed by executing a context-free feature-sharing algorithm.
15. The object/scene recognition system of claim 14 , wherein the recognition of the smaller-scale objects is performed by obtaining context information and classifying the smaller-scale objects based on the context information.
16. The object/scene recognition system of claim 15 , wherein the recognition of the complex objects is performed by executing a contextual bag of objects algorithm.
17. An object/scene recognition method comprising, with one or more mobile computing devices:
recognizing a plurality of visual features based on a map representation of a geo-spatial area of real world surroundings the one or more mobile computing devices generated based on temporal and spatial alignment of multi-dimensional image data obtained by one or more sensors, wherein the multi-dimensional image data includes at least one of two-dimensional image data and three-dimensional image data,
wherein the recognizing of the plurality of visual features includes recognizing of larger-scale objects, recognizing of smaller-scale objects by performing context-free object identification and contextual object identification, and recognizing of a complex object comprising a plurality of the smaller-scale objects.
18. The object/scene recognition method of claim 17 , further comprising:
temporally and spatially aligning the multi-dimensional image data; and
generating the map representation of the geo-spatial area of the real world surroundings of the mobile computing device based on the temporally and spatially aligned multi-dimensional image data.
19. The object/scene recognition method of claim 17 , wherein the recognizing of the complex object is performed using a classifier.
20. The object/scene recognition method of claim 17 , wherein the recognizing of the larger-scale objects is performed by determining a contextual frame of reference and using the contextual frame of reference to identify the larger-scale objects.
21. The object/scene recognition method of claim 17 , wherein the recognizing of the larger-scale objects is performed by executing an invariant multi-dimensional feature detection algorithm on data from at least one of the image sensors.
22. The object/scene recognition method of claim 17 , wherein the recognizing of the smaller-scale objects is performed by executing a context-free feature-sharing algorithm.
23. The object/scene recognition method of claim 22 , wherein the recognizing of the smaller-scale objects is performed by obtaining context information and classifying the smaller-scale objects based on the context information.
24. The object/scene recognition method of claim 23 , wherein the recognition of the complex objects is performed by executing a contextual bag of objects algorithm.