Edge Computing Object Identification via Segmentation and Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing distributed object identification systems face instability and high costs due to the need for multiple servers, high network bandwidth requirements, and security vulnerabilities, particularly when one component fails, leading to system unavailability.
Innovation Solution
A distributed object identification system utilizing edge computing devices that elect a central control device to monitor and respond to faults, perform object identification, and encrypt video streams, potentially integrated with a blockchain network for enhanced security and data integrity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If a distributed system is constructed based on computing clusters with multiple servers, then the processing capability is improved, but the system stability deteriorates because a fault in one component leads to system unavailability
Solution Approach 1:
The patent divides the distributed system into independent edge computing devices that each perform object identification locally. Each device operates autonomously without requiring centralized coordination, so that a fault in one device does not propagate to affect other devices. This segmentation into independent operational units resolves the contradiction by maintaining system stability while preserving processing capability through parallel local operations.
2Ease of operation
If video streams are uniformly transmitted to a computing cluster, then centralized processing is achieved, but network bandwidth requirements increase and security vulnerabilities arise
Solution Approach 1:
The patent extracts the video stream transmission requirement by performing object identification directly at the edge devices where the video streams originate. Instead of extracting and transmitting entire video streams to a central server, the system processes video data locally and only transmits identification results. This extraction approach dramatically reduces network bandwidth consumption while maintaining centralized-like coordination through result aggregation.
3Productivity
If multiple servers are deployed for distributed processing, then processing capacity is improved, but system cost increases
Solution Approach 1:
The patent enables edge computing devices to perform object identification independently using their own local resources and capabilities. Each device serves itself by maintaining local object identification models and processing video streams without requiring constant communication with or resources from other servers. This self-service approach allows the system to scale processing capacity by adding individual edge devices rather than requiring expensive centralized server infrastructure.
Data Source
AI summary
An edge computing device is provided including at least one memory configured to store computer program code and at least one processor configured to access said computer program code and operate as instructed by said computer program code. The edge computing device is included in a distributed object identification system, which includes a plurality of edge computing devices, and the edge computing device is determined as a central control device based on election from the plurality of edge computing devices. The computer program code includes first capturing code configured to cause the at least one processor to capture a video stream of an environment and first obtaining code configured to cause the at least one processor to obtain identity information of an object in the video stream by performing object identification on the video stream.