Smart Camera Edge Processing for Private Video Object Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing smart cameras require cloud-based processing and storage, raising privacy concerns and necessitating specialized hardware for autonomous operation without internet connectivity.

Innovation Solution

A device for local edge processing of video streams using a processor that performs object detection independently of cloud architecture, with scene detection engines running in parallel to analyze frames from multiple cameras, allowing for object identification and notification without relying on cloud services.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Power

If cloud-based processing is used for object detection, then processing power and storage capacity are improved, but privacy concerns increase and hardware requirements are reduced

Engineering Contradiction:

Improveprocessing powerVSAvoidprivacy concerns

Core Design Contradiction:

PowerVSObject-affected harmful factors

Solution Approach 1:

The patent extracts the object detection processing function from the cloud environment and relocates it to local edge devices. The edge device runs scene detection engines and object detection engines locally, processing video streams without transmitting them to the cloud, thereby maintaining processing capability while eliminating privacy concerns associated with cloud-based processing

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces an edge device as an intermediary between cameras and cloud services. This edge device performs local processing of video streams, acting as a mediator that reduces the need for cloud-based processing while maintaining system functionality. The edge device can operate independently or in conjunction with cloud services, providing flexibility in architecture

Inventive Principle:

Principle #24Intermediary (Mediator)

2Power

If cloud-based processing is used for object detection, then processing capability is improved, but hardware requirements are reduced

Engineering Contradiction:

Improveprocessing capabilityVSAvoidhardware requirements

Core Design Contradiction:

PowerVSDevice complexity

Solution Approach 1:

The patent segments the video processing workflow into multiple independent components: scene detection engines that perform preliminary analysis, object detection engines that perform detailed analysis, and frame queuing mechanisms. This segmentation allows each component to be optimized independently and processed in parallel, distributing computational load across multiple smaller units rather than requiring a single powerful cloud system

Inventive Principle:

Principle #1Segmentation

3Object-affected harmful factors

If autonomous operation without internet connectivity is required, then privacy is preserved, but hardware power requirements increase significantly

Engineering Contradiction:

Improveprivacy preservationVSAvoidhardware power requirements

Core Design Contradiction:

Object-affected harmful factorsVSUse of energy by moving object

Solution Approach 1:

The patent implements periodic action through frame sampling and selective processing. The scene detection engine analyzes frames at intervals rather than continuously, and the object detection engine processes only those frames that contain detected scenes. This periodic processing approach maintains autonomous operation capability while significantly reducing power consumption compared to continuous full-frame analysis

Inventive Principle:

Principle #19Periodic action

Solution Approach 2:

The patent applies partial action by having the scene detection engine perform preliminary filtering on video frames before passing them to the more computationally intensive object detection engine. Only frames that meet certain criteria (containing detected scenes) are queued for full object detection analysis, reducing the overall computational burden and power requirements while maintaining detection effectiveness

Inventive Principle:

Principle #16Partial or excessive action

4Productivity

If multiple video streams are processed in parallel, then detection coverage is improved, but processing complexity increases

Engineering Contradiction:

Improvedetection coverageVSAvoidprocessing complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the processing of multiple video streams by dedicating separate scene detection engines to each stream while sharing object detection resources. Each camera feed is processed independently through its own scene detection engine, allowing parallel processing of multiple streams. The segmented approach maintains detection coverage across all streams while managing complexity through modular architecture

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent merges the object detection resources across multiple video streams, allowing a single object detection engine to process frames from multiple scene detection engines. This combining approach reduces redundancy and processing complexity by sharing computational resources while maintaining the ability to handle multiple parallel video streams effectively

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS12412391B2System, device and method for object detection in video feeds

Publication Date: 2025.09.09 MONITOREAL LTD

AI summary

Embodiments are directed to a smart camera device that analyzes independent video streams.