For operators of large video camera networks it is a great challenge to ensure the image quality and the constant availability of all cameras in order to provide the required services.
Automated monitoring of the network connection of the cameras is considered to be state-of-the-art, but the quality control of the image by humans (operators) is usually not carried out due to the immense number of cameras and the associated high workload and costs. Inadequate image quality (or problems) of a camera are detected by the operator only in his monitoring tasks when explicitly seen by the user. This means that the actual task cannot be performed.
With the newly developed AViMon system, it is now possible to assess the image quality of several thousand cameras continuously and robustly and to automatically report cameras with problematic image quality or changed field of view.
Operators of large camera networks in public transport (road, rail, airport, ship), security (police, private security, cities) and industry can use AViMon to dramatically reduce the burden of detecting image defects. In addition, creeping deterioration of the image (aging, pollution ...) can be detected and measures can be taken immediately to prevent a total failure and thus ensure the highest possible availability of safety-critical infrastructures.
Ensuring image quality in video camera networks is essential for using the image content. Both when used by persons, e.g. to clarify security-related problems in security centers as well as when used by machines, e.g. as in the automated detection of traffic flow. With poor image quality or complete failure, a maintenance process must be initiated immediately to correct the problem. Due to the high number of video cameras and the sporadic use in normal operation, an ongoing assessment of image quality by humans is not feasible due to the extremely high workload and related costs.
Particularly outdoors, video cameras and the network infrastructure for signal transmission are subject to various influences, such as weathering, aging, malfunctions (electronic problems) or network anomalies. These can lead to serious image disturbances or breakdown, which make the intended use of the images impossible. Transmission and camera problems lead to image losses in the form of block or line dropouts, de-focused, low-contrast, over-exposed or completely lost images. Contamination, water or snow limit the field of view and unintendedly twisted cameras result in altered image content.
AViMon is a system that receives video streams or files from thousands of cameras in a rolling manner and automatically assesses the image quality. Only cameras with poor image quality are automatically reported by the system to a specialist who verifies the defect and, if necessary, immediately initiates maintenance and troubleshooting measures. By automating the image quality assessment, the time a person is involved in, and thus the cost, is reduced to a minimum.
AViMon is a robust and efficient camera surveillance system that is modular and highly scalable. It can be installed as a service in the local camera network on premise or in the future as a cloud service.
AViMon provides detailed information about the following image quality deficiencies:
- Image loss in the form of image blocks or lines
- Black images, low-contrast and over-exposed images
- De-focused / blurred images
- Flickering, unsteady, noisy video
- Sustainably changed image (through dirt, water, snow)
- Shifted image (by twisted camera)
- …and more
The AViMon team brings together more than 20 years of experience in research, development and user knowledge. Based on an existing solution implemented by Siemens for operating the ASFiNAG video network and JOANNEUM RESEARCH's market-leading video quality assessment tools, a scalable and robust image quality detection system for large video networks has been created. This solution is targeted to become the market standard in video quality assurance for local and cloud-based camera network services, meeting the fast-growing demand.
AViMon was supported by the research funding program "Mobility of the Future: Transport Infrastructure Research R&D Services (vif2016)“ of the Federal Ministry of Transport, Innovation and Technology, the Austrian Research Promotion Agency and the Autobahnen- und Schnellstraßen-Finanzierung-Aktiengesellschaft (ASFiNAG).