Edge Camera Hyperzoom Tracking for Low-Bandwidth Traffic Analytics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional security systems using conventional cameras often malfunction due to issues like camera failure, Wi-Fi connection loss, night-time operation failures, and inefficiencies caused by large image file sizes and network latencies.
Innovation Solution
The system employs a computer vision processor in cameras to generate hyperzooms for persons or vehicles, track traffic patterns, and perform analytics on the edge without relying on network usage. This includes using a Kalman Filter for position prediction, mobile semantic segmentation models for attribute extraction, and a key-value database for storing metadata.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional security cameras transmit video signals through Wi-Fi network, then remote viewing capability is provided, but network latency and large image file sizes cause inefficiency
Solution Approach 1:
The patent segments the video processing workflow into edge computing (local analytics, object detection, tracking) and cloud computing (historical analysis, complex queries). This segmentation allows real-time local processing to maintain efficiency while periodic cloud synchronization provides remote access capability.
Solution Approach 2:
The patent introduces an intermediary edge device (camera with embedded processor) that acts as a mediator between the video capture source and the remote user. This intermediary performs local analytics and buffers data, reducing network traffic and latency while maintaining remote accessibility.
2Loss of time
If security cameras continuously transmit video feeds, then real-time monitoring is achieved, but network bandwidth consumption increases
Solution Approach 1:
The patent extracts only the essential information (object detection results, tracking data, analytics) from the full video feed and transmits this compressed data to the cloud. This extraction approach maintains real-time monitoring capability while dramatically reducing network bandwidth consumption.
Solution Approach 2:
The patent changes the data transmission parameter from full-resolution video frames to compressed analytics data (JSON objects containing object coordinates, types, and tracking information). This parameter change reduces data size by orders of magnitude while preserving real-time monitoring effectiveness.
3Power
If camera processing is performed on cloud servers, then computational power is sufficient, but network dependency increases and latency occurs
Solution Approach 1:
The patent inverts the traditional cloud-centric processing model by placing the processing unit directly in the camera device. This inversion allows the camera to perform local analytics independently, reducing network dependency while maintaining sufficient computational power through embedded processors.
Solution Approach 2:
The patent enables the camera device to serve itself by performing local object detection, tracking, and analytics without requiring constant cloud server intervention. The camera autonomously processes video data and only synchronizes with the cloud when necessary, improving reliability and reducing latency.
Data Source
AI summary
A computer vision processor of a camera generates hyperzooms for persons or vehicles from image frames captured by the camera. The hyperzooms include a first hyperzoom associated with the persons or vehicles. The computer vision processor tracks traffic patterns of the persons or vehicles while obviating network usage by the camera by predicting positions of the persons or vehicles using a Kalman Filter from the first hyperzoom. The persons or vehicles are detected in the second hyperzoom. The positions of the persons or vehicles are updated based on detecting the persons or vehicles in the second hyperzoom. The first hyperzoom is removed from the camera. Tracks of the persons or vehicles are generated based on the updated positions. The second hyperzoom is removed from the camera. Track metadata is generated from the tracks for storing in a key-value database located on a non-transitory computer-readable storage medium of the camera.