GaugeCam Blog

The GaugeCam team met with Tom D. at NCSU yesterday to try to figure out how to commercialize our idea.  We had a pretty good brainstorming session.  All of us left the meeting with something to do so we can report back next week.  It is not like I do not have enough stuff to do already, but actually having something on the market in the next few months is a very fun thing to ponder.  We really do think we have found a niche.  I will post on GaugeCam and here as soon as we have things a little more carefull defined, but the first product will involve the ability to easily place cameras in very remote places where there is no hard wired nor close wireless connectivity nor power available, then receive a reliable stream of images displayed in a variety of user defined ways on a web stie.  Later, we will add some machine vision measurement capabilities to our setup, but for now, all we want to do is capture images for manual review.

Irene’s eye went right over our site! Unfortunately, some rain water managed to get into the casing of our prototype camera. It has been sent back to Colorado Video for damage assessment. We are confident that things should be OK though as there was water just at the bottom of the camera. Damage is probably moisture related so it is not as dramatic as if the camera had been ’swimming’…! In the mean time, the second version of our prototype is being shipped and will be installed on site as soon as possible.

Hurricane Irene is scheduled to pass through North Carolina in the next day or so.  GaugeCam is located in Raleigh.  Raleigh is expected to receive one to two inches of rain and winds in the range of 30-40 mph.  The tidal marsh camera could get hit hard.  You can see from François’s previous post that it is directly in the hurricane’s path and could experience up to a foot of rain and 100+ mph winds.  François sent his team out to the marsh to temporarily remove the camera.  Thanks for the help, guys!

CORRECTION!:  Troy just popped me a note to let me know the team has MOVED the camera to a higher location in the marsh where it will weather the storm.  It will be interesting to see how all this works out.

Marsh pictures have been streaming since July 17, 2011 with generally great success and we have learned a lot about what it takes to send images from a totally remote place, located half an hour away from the first paved road.  Images are streaming from the North River coastal marsh restoration project located near Beaufort, NC.

So, what does it take to have a GaugeCam system to measure water level in a stream?

First, it takes the camera that will do the trick of capturing pictures on a desired interval and of automatically sending the pictures to a server.  Second, it takes power: good batteries and good large solar panels.  Third, it takes a good white background with embedded fiducials. Fourth, it takes a stable camera mount.

First, the camera.  The GaugeCam project and company would probably have remained at the stage of a great potential project, had we not discovered the Colorado Video company and its very good products.  We are using a prototype of the Lookout V camera.

To be short, the camera has an embedded cell phone which sends pictures via a 3G card.  So in essence, one needs 3G network coverage for this camera to work properly.  We have experienced blackouts where the pictures do not go through.  The first reason for that was that our (still small but that is about to change) server was full…

There also seems to be a somewhat erratic pattern of missing pictures between 22:00 and 02:00 at night.  It seems to come and go and we have not been able to find a true reason on our side.  We are thinking that it might result from the 3G network operator, although we have no actual data to prove this.

Second, the power system.  We have a 55W solar panel to recharge two regular car batteries installed in parallel.  Voltage has never gone below 12.4V.  The marsh has no shade and is a great place for solar panels!

Third, the background.  This is a critical part of the set up as it has to withstand the rigors of the field (sun, water, biological activities, etc.), remain flat and vertical.  Actually, verticality is not necessary but it certainly helps to keep fouling to a minimum.  The background that you see in the pictures is a film of plastic that has been laminated over a ½ inch think piece of Plexiglas by professionals.  The background is attached to two posts that have braces to maintain verticality.  It is essential that the background does not move!  We will check regularly to verify that it does not!

Once the background is installed, it is important to measure the real world coordinates of the fiducials.  This is really the critical part that all hydrologists and equipment installers have to do.  Despite all efforts, it is our opinion that it is very difficult to make this measurement better than ±2 mm.  This is particularly difficult in the marsh where the water level was changing all the time because of the tides.  When the level hit the center of the lowest fiducial centers, water level was simultaneously measured above our reference value using a graduated ruler.  All future measurements are then dependent upon that first call…  Actually, other manual measurements will be done and if a bias towards over estimation or underestimation is detected, it will be taken as an offset to calibrate our instrument.

Fourth, it takes a good camera mount.  This might end up being one of the most challenging task, especially in remote areas.  Several options are possible: concrete, wooden or metal mounts…  We abandoned concrete as it would not be desirable there for many reasons.  We thought that metal would be very challenging because of varying moisture, salinity and temperatures.  So we went for wood (see pictures).  Braces were installed with the post to reduce vibrations from the wind.  However, we have already noticed that images are looking more and more to the right as the whole mounting device is warping.

The camera itself is protected in an open wooden box which size has been reduced to a minimum to minimize wind effect.

We are learning a lot from this experience and are adjusting our software and hardware accordingly!  We will keep you posted!

The metrology of the GaugeCam system was presented on Tuesday morning of the 2011 ASABE International Conference. The presentation featured images from our field cam and our lab camera, but the emphasis was on optimal system performance in the lab. We won’t present the specific findings here since we intend to publish those formally. However, we are pleased with the lab results and feel that we’ve found a “sweet spot” for camera location, angle and lens.

Since the conference had a strong academic presence, we really focused on the fine details of our lab analysis. We pointed out several sources of error that we encountered during our experimentation. Some of these sources we know how to correct (ex. glare from IR lighting), while others (ex. image distortion, unless you want to perform a nonlinear model of the image) must simply be avoided when applying this technology. Much of this material is basic knowledge to the machine vision community, but it is important to discuss with hydrologists who may want to apply the technology.

Yes, we can.

We were pleased with the turnout for the presentation, which included attendees from government agencies and universities, both domestic and international. The audience also asked some good questions about system maintenance and whether our results rival other measurement techniques.

On the official ASABE program, the contributors to this presentation were Troy Gilmore, François Birgand, Ken Chapman and Andrew Brown. Kelly Chapman contributed significantly as a lab volunteer this summer and she is currently finalizing our documentation so we can release the version of GRIME used in the study. Christian Chapman set up the original GaugeCam server. Randall Etheridge and Brad Smith worked to set up the salt marsh camera, which gave us access to some nice demonstration images. Thanks to everyone for their hard work. We look forward to seeing our results in print!

All the equipment is up and running at the ASABE Conference in Louisville, Kentucky.  We experienced the usual conference equipment and software hiccups, but Andrew, Troy, and François worked it all out.  Everything is up and running as planned and the GaugeCam booth is experiencing a continuous flow of interested conference attendees.

GaugeCam’s demonstrates technologies developed to measure water levels in the wild for the BAE lab at North Carolina State University.  We show live images arriving at the GaugeCam booth from a solar powered, cell-phone equipped Colorado Video camera of a water scene in a coastal North Carolina marsh.  The results are presented in real-time on the internet as a graph of water levels coupled with archived images.  There is a second live demo that features images arriving at a server in the booth from a Microseven camera pointed at a water column in the booth and presented, again, as a graph of the water level measurements coupled with an image archive.

GaugeCam’s purpose at this conference is to present technology that retrieves images and sensor data from remote sites, presents both raw and processed data automatically on the web and allows the user to control motors, weirs, lights, and other actuators either manually or based on sensor conditions and user specified logic via the internet.  The feedback we receive from conference attendees will help us refine our current commercial offerings, identify consulting opportunities, and possible even identify new product offerings.

Still to come:  Troy will present the results of the NCSU research based on data gathered with GaugeCam equipment.

The GaugeCam team has been putting in long days for the last couple of weeks to prepare the presentation Troy will deliver and the demonstration of our water column at the American Society of Biological and Agricultural Engineers (ASABE) conference in Louisville, KY on August 7-10, 2011. We will plan to show a version of the laboratory column used at our NCSU laboratory to demonstrate the capabilities of our remote camera and the image evaluation software to measure water height in the wild.  Troy will describe our methods and our most important results in his presentation.  Anyone who would like to get a copy of our desktop software is welcome to come by our booth to see the software in operation and talk to Andrew, Francois, and Troy.  We were hoping to be able to hand out copies of the software at the conference, but are making last minute “ease of use” adjustments to the software and documentation so it will be more usable by other researchers.  We plan to make the software available for download b August 19.  Leave your email or snail mail with us (or just drop us a line here) and we will let you know where to go on the web as soon as the alpha release version of the software is available.

We are looking forward to seeing you there!

GaugeCam will be presented at the ASABE International Conference in Louisville, KY on Aug 7-10, 2011.

GRIME has been put to the test lately. We’ve been processing loads of images and GRIME is performing well! We’re looking forward to compiling all of this data and really assessing our performance.

Imaging Source, one of our camera providers has put up an article I wrote on our work at GaugeCam.  It is worth checking out.