Aligned time surfaces via dense optical flows for event-based object detection, identification and tracking
Systems and methods are provided for producing aligned and aggregated aligned time surfaces. An exemplary method includes: receiving camera event data from a first location; computing, using the camera event data, a plurality of unaligned time surfaces; determining a plurality of optical flows corresponding to the unaligned time surfaces; determining, using the optical flows, an accumulated optical flow for mapping displacement between any two time periods; producing, using the optical flows, for a plurality of consecutive mapped time surfaces; performing, using the accumulated optical flow, warping of the mapped time surfaces producing corresponding aligned time surfaces; and aggregating the aligned time surfaces to produce an image having lowered blurring of object boundaries.
1 . A system for producing aligned time surfaces, the system comprising one or more computing device processors; and
one or more computing device memories, coupled to the one or more computing device processors, the one or more computing device memories storing instructions executed by the one or more computing device processors, wherein the instructions are configured to:
receive camera event data from a first location;
compute, using the camera event data, a plurality of unaligned time surfaces;
determine a plurality of optical flows corresponding to the unaligned time surfaces;
determine, using the optical flows, corresponding displacement parameters by respectively combining (a) a location to be mapped and (b) a result of using an integral on an optical flow of the optical flows for mapping displacement between any two time periods;
produce, using the displacement parameters, a plurality of consecutive mapped time surfaces;
perform warping, using the optical flows, of the mapped time surfaces producing corresponding aligned time surfaces; and
aggregate the aligned time surfaces to produce an image having lowered blurring of object boundaries.
2 . The system of claim 1 , wherein the first location is an event camera system.
3 . The system of claim 1 , wherein the first location is a remote storage device for storing the camera event data.
4 . The system of claim 1 , wherein while computing the unaligned time surfaces, the instructions are configured to aggregate the unaligned time surfaces in accordance with time and space.
5 . The system of claim 1 , wherein the unaligned time surfaces comprise information of at least one moving object.
6 . The system of claim 1 , wherein while determining the plurality of optical flows, the instructions are configured to determine a velocity of at least one moving object in the unaligned time surfaces.
7 . The system of claim 1 , wherein while performing warping of the unaligned time surfaces, the instructions are configured to align, using an accumulated optical flow, the aligned time surfaces in accordance with time and space.
8 . The system of claim 1 , wherein the instructions further configured to send the image downstream to an object identification, detection, and/or tracking process.
9 . The system of claim 8 , wherein while sending the image downstream, the instructions are configured to execute the object identification, detection, and/or tracking process on a remote system.
10 . A method for producing aligned time surfaces, the method comprising:
receiving camera event data from a first location;
computing, using the camera event data, a plurality of unaligned time surfaces;
determining a plurality of optical flows corresponding to the unaligned time surfaces;
determining, using the optical flows, corresponding displacement parameters by respectively combining (a) a location to be mapped and (b) a result of using an integral on an optical flow of the optical flows for mapping displacement between any two time periods;
producing, using the displacement parameters, a plurality of consecutive mapped time surfaces;
performing, using the optical flows, warping of the mapped time surfaces producing corresponding aligned time surfaces; and
aggregating the aligned time surfaces to produce an image having lowered blurring of object boundaries.
11 . The method of claim 10 , wherein the first location is an event camera system.
12 . The method of claim 10 , wherein the first location is a remote storage device for storing the camera event data.
13 . The method of claim 10 , wherein the computing the unaligned time surfaces comprises aggregating the unaligned time surfaces in accordance with time and space.
14 . The method of claim 10 , wherein the unaligned time surfaces comprise information of at least one moving object.
15 . The method of claim 10 , wherein determining the plurality of optical flows comprises determining a velocity of at least one moving object in the unaligned time surfaces.
16 . The method of claim 15 , wherein an accumulated optical flow comprises accumulated information associated with processing of the optical flows.
17 . The method of claim 16 , wherein performing warping of the unaligned time surfaces comprises aligning, using the accumulated optical flow, the aligned time surfaces in accordance with time and space.
18 . The method of claim 10 further comprising sending the image downstream to an object identification, detection, and/or tracking process.
19 . The method of claim 18 , wherein sending the image downstream comprises executing the object identification, detection, and/or tracking process on a remote system.
20 . A non-transitory computer-readable storage medium storing instructions which when executed by a computer cause the computer to perform a method for producing aligned time surfaces, the method comprising:
receiving camera event data from a first location;
computing, using the camera event data, a plurality of unaligned time surfaces;
determining a plurality of optical flows corresponding to the unaligned time surfaces;
determining, using the optical flows, corresponding displacement parameters by respectively combining (a) a location to be mapped and (b) a result of using an integral on an optical flow of the optical flows for mapping displacement between any two time periods;
producing, using the displacement parameters, a plurality of consecutive mapped time surfaces;
performing warping, using the optical flows, of the mapped time surfaces producing corresponding aligned time surfaces; and
aggregating the aligned time surfaces to produce an image having lowered blurring of object boundaries.