Edge Learning Classifier for Incident Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Public safety personnel often fail to activate image capture devices in time to recognize objects of interest during incidents, due to unexpected struggles or removal of the device, limiting the accuracy of object detection and response efforts.
Innovation Solution
A portable electronic device equipped with multiple cameras and sensors that automatically detects incidents, selects the appropriate camera, captures images, and initiates an edge learning process to create classifiers for identifying subjects of interest, which are then transmitted to nearby devices for recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If public safety personnel manually activate image capture devices, then the device can capture images for object detection, but the personnel may not be able to activate the camera in time due to unexpected struggles or removal of the device
Solution Approach 1:
The image capture device automatically activates and captures images without requiring manual activation by public safety personnel. The device monitors its own operational state and autonomously initiates image capture when incidents are detected, eliminating the time loss associated with manual activation while ensuring accurate object detection through continuous automated operation.
2Speed
If edge learning process is performed locally on the portable electronic device, then real-time object recognition can be achieved, but the device complexity and computational resource requirements increase
Solution Approach 1:
The learning system is divided into two segments: a lightweight edge learning component that runs locally on the portable electronic device for real-time object recognition, and a more comprehensive cloud-based learning system that performs extensive training and model updates. This segmentation allows the device to achieve fast local recognition without bearing the full complexity of the complete learning system.
Solution Approach 2:
A simplified edge learning model acts as an intermediary between the image capture device and the full learning system. This intermediary performs initial object recognition locally with reduced complexity, then transmits relevant data to the cloud-based system for further processing and model improvement, balancing speed and complexity requirements.
3Reliability
If multiple cameras are used to capture images during incidents, then the likelihood of capturing the subject of interest increases, but the device complexity and data processing requirements increase
Solution Approach 1:
Multiple cameras are pre-positioned on the device to capture images from different angles and perspectives. During incidents, the system automatically selects and activates the appropriate camera based on the incident type and orientation, having already prepared the multi-camera configuration in advance. This preliminary setup increases capture reliability without requiring complex real-time decision-making about camera selection.
Data Source
AI summary
A portable electronic device and method. The portable electronic device includes a first camera, a second camera, an electronic processor, and one or more sensors. The electronic processor is configured to detect, based on information obtained from the one or more sensors, an incident and select a camera responsive to the incident. The electronic processor is further configured to capture an image using the selected camera and determine, within the image, a subject of interest, wherein the subject of interest is at least one selected from the group consisting of a person, an object, and an entity. The electronic processor is also configured to initiate an edge learning process on the subject of interest to create a classifier for use in identifying the subject of interest and transmit the classifier to a second portable electronic device within a predetermined distance from the portable electronic device.