Fisheye Surveillance System Audio-Triggered Image Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surveillance systems, particularly those using Pan-Tilt-Zoom (PTZ) cameras, require manual control and lack automatic triggering capabilities, leading to delays in capturing and responding to events at locations of interest, which can result in missing critical incidents.
Innovation Solution
An active audio-visual system comprising a Fisheye image capturing device, an audio module, and a computing module that uses Direction of Arrival (DOA) data from audio signals to automatically generate high-resolution perspective images, leveraging a fisheye correction scheme and single-image-super-resolution techniques to cover large areas and capture events promptly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If PTZ cameras are used to monitor open areas, then the coverage area is improved, but the response time deteriorates due to manual control requirements
Solution Approach 1:
The system performs preliminary actions by continuously capturing wide-angle images with the fisheye camera and pre-processing them into non-distorted perspective images before events occur. When an audio event is detected, the system can immediately retrieve and display the pre-processed image corresponding to the event location, eliminating the time delay associated with manual PTZ camera operation.
Solution Approach 2:
The invention replaces the mechanical manual control system of PTZ cameras with an automated audio-visual system. The audio module detects events and automatically triggers the display of corresponding pre-processed images, substituting the mechanical pan-tilt-zoom operations with an automated computational approach that retrieves images based on audio event location.
2Manufacturing precision
If fisheye correction scheme is applied, then the image quality is improved, but the processing complexity increases
Solution Approach 1:
The system applies fisheye correction in advance to all captured wide-angle images, converting them into non-distorted perspective images before events occur. This pre-processing approach ensures that when an audio event is detected, the system can immediately retrieve the already-corrected image without needing to perform real-time fisheye correction, thus maintaining high image quality while avoiding complex real-time processing.
3Measurement precision
If single-image-super-resolution is used, then the image resolution is improved, but the computational load increases
Solution Approach 1:
The system applies single-image-super-resolution to all captured wide-angle images in advance, generating high-resolution perspective images before events occur. This pre-processing approach ensures that when an audio event is detected, the system can immediately retrieve the already-enhanced high-resolution image without needing to perform real-time super-resolution processing, thus maintaining high image resolution while reducing real-time computational load.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system enables rapid and efficient capture of high-resolution images of events before they disappear from view, providing timely and comprehensive coverage of locations of interest without manual intervention.
Implementation Method 1
an audio module provided at the location-of-interest, the module configured to capture audio signals and to determine a Direction of Arrival (DOA) of the audio signals
Implementation Method 2
the device configured to continuously capture wide-angle image frames and to convert the captured wide-angle image frames into non-distorted perspective images using a set of calibration parameters generated by a fisheye correction scheme
Implementation Method 3
instruct the Fisheye image capturing device to generate, using a trained single-image-super-resolution (SISR) scheme, high-resolution perspective images based on the non-distorted perspective images
Data Source
AI summary
An active audio-visual system and method for identifying events occurring at a location-of-interest is provided. In particular, a system that comprises a Fisheye image capturing device, an audio module and a computing module communicatively coupled to both the image capturing device and the audio module whereby all the modules and devices are provided at the location-of-interest is provided. In operation, the audio module is configured to issue a notification to the computing module when an audio event is detected at the location-of-interest. The notification, which will contain a point of origin of the audio event, will then be used by the computing module to cause the calibrated Fisheye image capturing device to provide a captured high-resolution perspective image of the point of origin of the audio signal to an event identification module for further processing.