Low light surveillance system with dual video streams
Systems and methods for a low light surveillance system with dual video streams are described. A video camera may include an image sensor and an infrared blocking filter. Responsive to detecting low light conditions, the infrared blocking filter may be disabled and the full spectrum video data from the camera may be sent in one video stream that is processed for user display, such as rendering in gray scale, and another video stream that is used for analytics processing, such as event detection.
1. A system, comprising:
a video camera comprising:
an image sensor; and
an infrared blocking filter; and
a controller, comprising a first memory and a first processor, configured to:
determine a low light condition;
disable, responsive to determining the low light condition, the infrared blocking filter;
send a first video stream from the image sensor to a user device for displaying the first video stream; and
send a second video stream from the image sensor to an analytics engine, wherein:
the analytics engine is stored in a second memory for execution by a second processor configured for processing, using a first analytical model, the second video stream; and
the first analytical model is configured to detect low light events from a full spectrum video stream from the image sensor.
2. The system of claim 1 , wherein the controller is further configured to process, prior to sending the first video stream to the user device, the first video stream into a gray scale video stream.
3. The system of claim 1 , further comprising the user device, wherein the user device comprises:
a third processor;
a third memory;
a graphical user display configured for selectively displaying the first video stream to a user of the user device; and
an input device configured for selectively navigating the first video stream on the graphical user display.
4. The system of claim 1 , further comprising the analytics engine, wherein the analytics engine is further configured to use
a second analytical model to detect normal light events from an infrared filtered video stream.
5. The system of claim 4 , wherein:
the analytics engine is further configured to send, responsive to detected low light events, low light event parameters to the controller; and
the controller is further configured to map the low light event parameters to the first video stream.
6. The system of claim 5 , further comprising the user device, wherein:
the user device comprises a third processor and a third memory; and
the user device is configured to:
receive the first video stream and the low light event parameters;
display the first video stream as a gray scale video on a graphical user display; and
overlay, based on the low light event parameters, at least one low light event indicator on the gray scale video.
7. The system of claim 4 , wherein the analytics engine is further configured to:
determine a set of low light training data comprising pre-determined low light events in full spectrum video data;
iteratively train, using the set of low light training data, the first analytical model;
determine, based on the iterative training, a set of low light model weights for the first analytical model; and
apply the set of low light model weights to processing the second video stream.
8. The system of claim 7 , wherein the analytics engine is further configured to:
determine a set of normal light training data comprising pre-determined normal light events in infrared filtered video data;
iteratively train, using the set of normal light training data, the second analytical model;
determine, based on the iterative training, a set of normal light model weights for the second analytical model; and
apply, responsive to the controller determining a normal light condition, the set of normal light model weights to processing a normal light video stream from the image sensor.
9. The system of claim 1 , further comprising:
a video storage device, comprising a non-volatile storage medium, configured to store the first video stream in the non-volatile storage medium; and
the user device, wherein:
the user device comprises a third processor and a third memory; and
the user device is configured to:
selectively retrieve a portion of the first video stream from the video storage device; and
display the portion of the first video stream on a graphical user display of the user device.
10. The system of claim 9 , further comprising the analytics engine configured to detect low light events from the second video stream, wherein:
the video storage device is further configured to store the second video stream; and
the analytics engine is further configured to selectively retrieve the second video stream from the video storage device for processing.
11. A computer-implemented method, comprising:
determining a low light condition for a video camera, wherein the video camera comprises:
an image sensor; and
an infrared blocking filter;
disabling, responsive to determining the low light condition, the infrared blocking filter;
sending a first video stream from the image sensor to a user device; and
detecting, using a first analytical model in an analytics engine, low light events from a second video stream, wherein the second video stream is a full spectrum video stream from the image sensor.
12. The computer-implemented method of claim 11 , further comprising:
processing, prior to sending the first video stream to the user device, the first video stream into a gray scale video stream.
13. The computer-implemented method of claim 11 , further comprising:
displaying, on a graphical user display of the user device, the first video stream to a user of the user device; and
selectively navigating, using an input device of the user device, the first video stream on the graphical user display.
14. The computer-implemented method of claim 11 , further comprising:
detecting, using a second analytical model in the analytics engine, normal light events from an infrared filtered video stream captured by the video camera during a normal light condition.
15. The computer-implemented method of claim 14 , further comprising:
mapping low light event parameters corresponding to the detected low light events from the second video stream to the first video stream.
16. The computer-implemented method of claim 15 , further comprising:
receiving, at the user device, the first video stream and the low light event parameters;
displaying, on a graphical user display of the user device, the first video stream as a gray scale video; and
overlaying, based on the low light event parameters, at least one low light event indicator on the gray scale video.
17. The computer-implemented method of claim 14 , further comprising:
determining a set of low light training data comprising pre-determined low light events in full spectrum video data;
iteratively training, using the set of low light training data, the first analytical model;
determining, based on the iterative training, a set of low light model weights for the first analytical model; and
applying the set of low light model weights to processing the second video stream.
18. The computer-implemented method of claim 17 , further comprising:
determining a set of normal light training data comprising pre-determined normal light events in infrared filtered video data;
iteratively training, using the set of normal light training data, the second analytical model;
determining, based on the iterative training, a set of normal light model weights for the second analytical model; and
applying, responsive to determining a normal light condition, the set of normal light model weights to processing a normal light video stream from the image sensor.
19. The computer-implemented method of claim 18 , further comprising:
storing the first video stream and the second video stream in a video storage device;
selectively retrieving, by the user device, a first portion of the first video stream from the video storage device;
displaying the first portion of the first video stream on a graphical user display of the user device;
selectively retrieving, by the analytics engine, a second portion of the second video stream from the video storage device for processing; and
detecting low light events from the second portion of the second video stream.
20. A storage system, comprising:
a video camera comprising:
an image sensor; and
an infrared blocking filter;
a processor;
a memory;
means for determining a low light condition for the video camera;
means for disabling, responsive to determining the low light condition, the infrared blocking filter;
means for displaying a first video stream from the image sensor on a graphical user display,
wherein the first video stream is a gray scale video stream; and
means for detecting low light events from a second video stream from the image sensor, wherein the means for displaying the first video stream overlays detected low light events from the second video stream on the first video stream.