Adaptive video end-to-end network with local abstraction
An adaptive video end-to-end network is described that uses local abstraction. One example includes an image sensor to generate a sequence of images, a processor coupled to the image sensor to analyze the sequence of images to detect an event, to select images related to the event and to generate metadata regarding the event, and a communications interface coupled to the processor to send the metadata information through a network connection to a central node.
1. An apparatus comprising:
a networked camera to send sparse image data and metadata to a central node so that the central node is able to perform detection, tracking and recognition on the received sparse image data and metadata;
the networked camera including a processor to abstract semantic forms, to compress data from the images in the feature domain using the semantic abstraction, to generate sparse image data describing low level features based on the compressed image data, to generate actionable insights based on the semantic abstraction, and to select the sparse image data and metadata to be sent to a central node based on the insights;
said processor to detect an object in one particular a category in a video and to extract a frame portion including an object in the one category from a frame of the video; and
a video encoder coupled to said processor device to encode at least one frame portion with the extracted object and less than all of the frame, and to send out the encoded frame portion of the video frame depicting an object in said category to a vision-based analysis system, the encoded frame portion including less than all the data of the frame in the video, and the vision-based analysis system, when coupled to said video encoder, to receive the encoded portion and to perform vision-based analysis on the portion.
2. The apparatus of claim 1 said processor to detect only one category of objects.
3. The apparatus of claim 2 wherein said category is one of a license plate and a human face.
4. The apparatus of claim 1 wherein said video encoder to perform spatial compression.
5. The apparatus of claim 1 including a server coupled to said video encoder to perform vision-based analysis.
6. The apparatus of claim 1 wherein the vision-based analysis system is coupled to said video encoding.
7. One or more non-transitory computer readable media storing instructions executed by a processor to perform a sequence comprising:
sending sparse image data and metadata to a central node so that the central node is able to perform detection, tracking and recognition on the received sparse image data and metadata;
abstracting semantic forms, to compress data from the images in the feature domain using the semantic abstraction;
generating sparse image data describing low level features based on the compressed image data;
generating actionable insights based on the semantic abstraction;
selecting the sparse image data and metadata to be sent to a central node based on the insights;
detecting an object in one particular a category in a video and to extract a frame portion including an object in the one category from a frame of the video;
encoding at least one frame portion with the extracted object and less than all of the frame; and
sending out the encoded frame portion of the video frame depicting an object in said category to a vision-based analysis system, the encoded frame portion including less than all the data of the frame in the video, and the vision-based analysis system, when coupled to said video encoder, to receive the encoded portion and to perform vision-based analysis on the portion.
8. The media of claim 7 , further storing instructions to detect only one category of objects.
9. The media of claim 8 , wherein said category is one of a license plate and a human face.
10. The media of claim 7 , further storing instructions to perform spatial compression.
11. The media of claim 7 , further storing instructions to send out data for one of detection, tracking and recognition.
12. The media of claim 7 , further storing instructions to perform vision-based analysis.
13. An apparatus comprising:
means for sending sparse image data and metadata to a central node so that the central node is able to perform detection, tracking and recognition on the received sparse image data and metadata;
means for abstracting semantic forms, compressing data from the images in the feature domain using the semantic abstraction, generating sparse image data describing low level features based on the compressed image data, generating actionable insights based on the semantic abstraction, and selecting the sparse image data and metadata to be sent to a central node based on the insights;
means for detecting an object in one particular a category in a video and to extract a frame portion including an object in the one category from a frame of the video; and
means for encoding at least one frame portion with the extracted object and less than all of the frame, and sending out the encoded frame portion of the video frame depicting an object in said category to a vision-based analysis system, the encoded frame portion including less than all the data of the frame in the video, and the vision-based analysis system, when coupled to said video encoder, to receive the encoded portion and to perform vision-based analysis on the portion.
14. The apparatus of claim 13 including a camera that detects only one category of objects.
15. The apparatus of claim 14 wherein said category is one of a license plate and a human face.
16. The apparatus of claim 13 wherein said video encoder to perform spatial compression.
17. The apparatus of claim 14 said camera to send out data for one of detection, tracking and recognition.
18. The apparatus of claim 13 including a server coupled to said video encoder to perform vision-based analysis.
19. The apparatus of claim 13 wherein the vision-based analysis system is coupled to said video encoding.