Revision #10Recommend a SolutionFork
Nairobi
Washington DC
London
Bangalore
Use webcam images/video to estimate water stage/flows in natural and manmade water channels/rivers. Should provide an inexpensive way to monitor river/canal flows. Currently high cost of flow sensors and telemetry (est. US$ 2000+) make it difficult to deploy in developing country settings and result in inadequate real-time knowledge of water flows. This information is critical for operational management (e.g. dam/canal operations) and for disaster management (e.g. flood early warning systems). Using simple and ubiquitous web cams (e.g. from fixed locations) to get images/video feeds (and sent using simple GSM telemetry or other options as appropriate) can provide information that can be interpreted (e.g. using pattern recognition software) to correlate with actual water flows (using short measurement records for calibration). This can then create a system where future flows can be interpreted directly from the webcam imagery at a fraction of the cost of modern sensors.
Recommended Solutions:
Example:
This can be used in several countries (e.g. Malawi in the Shire Basin, or in any other river system). Easily scalable, combined with lower-cost telemetry options to set up an integrated visualization system and inputs into real-time decision support systems for operational or flood management.
User Stories:
Use pattern recognition techniques and statistical correlation analysis.
Constraints:
Initial datasets for calibration, adequate training.
Extra Credit:
Combine with telemtery system, develop webcam integrated monitoring device with onsite flow calculator to minimize data to be telemetered.
Similar Projects and Resources:
Would be useful to have a location where flow measurement is already being undertaken and is available (e.g. existing USGS site) and have a short webcam footage from fixed location especially covering a range of flows. This can then be used to calibrate the pattern recognition to estimate the flows and then validated on another period.
Next Steps and Sustainability:
Yes -many projects (incl. World Bank financed) in countries (e.g. Malawi) and river basins around the world that can make use of such a solution.
Qualitative Impact:
It will help set up a relatively inexpensive water monitoring system that can help early warning or operational management of water infrastructure. This can potentially save many lives if used with adequate early warning systems.
Quantitative Impact:
Could be a solution that helps millions around the world impacted by floods routinely and millions more that depend on adequate and timely water resources management.
Problem Definition Category:
Event:
Comments
Brilliant idea. If someone
Anyone already thought about
Anyone already thought about using a Wii controller?
I found this articles - http://www.wired.com/wiredscience/2009/12/wiimote-science/
http://www.newscientist.com/blogs/shortsharpscience/2010/12/hacked-wiimo...
If not the Wii controler..
If not the Wii controler.. perhaps pattern recognition?
http://opencv.willowgarage.com/wiki/Welcome
There are several sites that
There are several sites that have webcams at gauge sites. Perhaps something like this may help: http://eathorvertonwebcam.org.uk/index2.html The challenge could be to use the historical imagery there to (i) read off the levels from Cam 1 and (ii) use pattern recognition on Cam 2 to estimate the levels from (i) at the same time. The archived images can be downloaded in advance.