Web   ·   Wiki   ·   Activities   ·   Blog   ·   Lists   ·   Chat   ·   Meeting   ·   Bugs   ·   Git   ·   Translate   ·   Archive   ·   People   ·   Donate

#publiclab, 2012-11-03

 « Previous day | Index | Today | Next day »     Channels | Search | Join

All times shown according to UTC.

Time Nick Message
02:04 cfastie has quit IRC
02:45 Shannon <Shannon!~Shannon@qwebirc.media.mit.edu> has joined #publiclab
02:46 Shannon #endmeeting
02:47 warren_ <warren_!~warren@qwebirc.media.mit.edu> has joined #publiclab
02:47 warren_ #endmeeting
02:48 warren_ has quit IRC
02:58 Shannon has quit IRC
14:06 meeting Meeting started Sat Nov  3 14:06:17 2012 UTC. The chair is warren. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:06 Useful Commands: #action #agreed #help #info #idea #link #topic #endmeeting
14:11 liz <liz!~liz@qwebirc.media.mit.edu> has joined #publiclab
14:11 liz #map stitching
14:13 crabi <crabi!~crabi@qwebirc.media.mit.edu> has joined #publiclab
14:22 crabi is anybody chatting here?
14:25 Hudonnoodles @crabi - we will be
14:25 we're having difficulties with power right now, so hopefully we'll be up and running in a bit
14:26 #topic DIY Map Making Overview
14:27 liz has quit IRC
14:28 mathew <mathew!~mathew@qwebirc.media.mit.edu> has joined #publiclab
14:29 Hudonnoodles Stewart Long: use this opportunity to show some tips on making maps and walk through different situations
14:29 after presentation we can discuss future projects we can work on as a group
14:30 Typically you go out and do a flight, and you have a folder of images on your machine. In this sample we only have 8.
14:30 Before I open the map program, I just look at the pictures and look at the overall flight
14:32 I use thumbnails
14:32 mapknitter.org
14:32 I log in with my usernap on mapknitter so projects can stay together under author
14:32 liz <liz!~liz@qwebirc.media.mit.edu> has joined #publiclab
14:32 Hudonnoodles start off by creating title and location information
14:32 liz @jen this chat says "no title set"
14:33 Hudonnoodles Really? I did a #topic DIY Map Making Overview at 10:28
14:33 @ Liz do we need to do anything else?
14:34 liz the top of my chat says "(no topic set)"
14:34 Hudonnoodles same here
14:34 liz hmmmm bug @jeff
14:34 Hudonnoodles @liz, it's apparently unrelated
14:35 you want to know what part of the flight you want to map
14:35 sometimes it can be at the highest altitude or a specific area.
14:36 the higher the altitude, the easier its going to process (or stitch) into a map
14:36 there's going to be distortion around the perimeter.
14:37 on the screen is a mapknitter project with 25 overlays
14:38 Today we want to make a map of Lumcon, so later today there will be a mapknitter project
14:38 We will be able to see it on the screen with better color
14:38 You take each picture, you fit it, and then you lock it in place
14:38 MapKnitter has this feature where you can make the entire image transparent or half opaque
14:39 once you're satisfied with the image, you lock it in place and make it transparent
14:39 you want to take each new picture and lock it again the base map
14:39 *against (not again)
14:39 You don't want to build upon your errors
14:39 It's better to always go against your picture of reference
14:39 When you're finally done with the whole scene, I make them all visible again and then export
14:39 It's all about doing one picture at a time
14:40 Fitting each picture on the map
14:40 Stewart continue to show examples
14:40 On maps like Tikehau Before/After, you use rocks and vegetation that's there year-to-year, rather than seasonal
14:40 In urban areas there are a lot of features, but in rural areas you have to look for a big tree or a big tree
14:41 something that is there year after year
14:41 Burning Man 2010
14:41 Over 300 pictures
14:41 there are lots and lots of pictures in it!
14:41 On Screen: Flying Over Black Rock City - Geodata Collection - Video
14:41 There were 16 site lines over the city
14:42 Build out the Burning Man 2010 image - there was no ground control, we just used yellow lines
14:42 Vector map came from the planning design from the city
14:43 USe higher altitude flight to use for your lower altitude flight (using large scale to create the small scale)
14:43 Jeff W: you can sometimes find a satellite map that's not in the google imagery
14:43 In the country of Georgia, we got a map from the country that was significantly better than Google Maps.
14:43 Bootstrapping process
14:44 Mathew: Do you see this as a long term problem, if people continue to stitch over our maps since they're sometimes more up-to-date?
14:44 Jeff W: It happens, but it's not that common now that our maps are sometimes Google Maps
14:45 Chris F: Gowanus Canal, the imagery is going to improve. It's going to be harder and harder to have a balloon map that's better than a Google Earth image
14:45 Stewart Long: MapKnitter takes number crunching away from user
14:45 On Screen: behind the scenes work, GDAL Utilities
14:46 gdal.org/gdal_utilities.html
14:46 On Screen: Mission Dolores Park, San Francisco
14:46 It's a rectangular park, we walked around the perimeter of the park.
14:47 It was a project that had to be done in Photoshop
14:47 Make the image in photoshop, then you go into Google Maps. Create ground points. Distributed evenly throughout where the mapping was done
14:47 Take 24 points and save it as a KML file
14:47 (so you can read the lat/long points)
14:48 GDAL commands, every single ground control point is taking the xy from Photoshop in the document space and marrying it with the lat/long in real world
14:48 It takes tme
14:48 Dan Beaver: Brain numbing!
14:49 *Beavers
14:49 liz gdal_translate -of GTiff -a_srs EPSG:4326
14:49 Hudonnoodles Stewart Long: Gdal2tiles
14:49 take picture and look at it different levels
14:49 liz gdalwarp -of GTiff -t_srs EPSG:4326
14:49 Hudonnoodles To do this manually, it takes 1-2 hours
14:49 MapKnitter takes that manual work away from the user
14:50 Chris Fastie: Is there a size limit in MapKnitter?
14:50 liz gdaltotiles.py --title="<2012....etc"
14:50 Hudonnoodles Jeff W: We used to have an unlimited number when we were cartagen knitter. We used to have spam exports.
14:50 But now we've fixed that. We do have limits, we have uLimits. All the processes have to run inside the "cage."
14:51 Limited to ram and disk space
14:51 Your entire job is taking up gigabytes of space. I don't remember the limit, but I think its 5 gigabytes
14:51 If there's more demand, we can raise your limit by request
14:51 liz detailed instructions in the following notes: http://publiclaboratory.org/no[…]large-map-project
14:52 http://publiclaboratory.org/wiki/gdal-commands
14:52 Hudonnoodles Jeff W: The first time we realized it was a problem when part of a map was in France and the other part was in New York City. Lots of pixels in between!
14:53 Stewart Long: Related maps, like Burning Man 2009 and Burning Man 2010
14:53 http://publiclaboratory.org/ma[…]n-2010/2010-09-04
14:53 Hudonnoodles has quit IRC
14:53 Hudonnoodles <Hudonnoodles!~Hudonnood@qwebirc.media.mit.edu> has joined #publiclab
14:54 Hudonnoodles National Map
14:54 viewer.nationalmap.gov/viewer/#
14:55 liz download geotiffs of orthoimagery for use in photoshop
14:55 Hudonnoodles And ask for help! If you don't know what to do.
14:55 Stewart Long: highlighting a few interesting projects
14:55 liz and National Map is "having intermittent issues" ;)
14:55 Hudonnoodles Ft. Morgan Tracks in Google Earth
14:56 http://www.flickr.com/photos/g[…]earth/4897001954/
14:56 Oblique, not straight down
14:56 liz "you're the 'lawnmower' when you walk around your site to get swaths of imagery"
14:57 Hudonnoodles Balloon Elevation Profile in Google Earth 5.2
14:57 http://www.flickr.com/photos/g[…]earth/4701646651/
14:57 Don: How big was the balloon?
14:58 Stewart: Standard Public Lab balloon
14:58 Lafayette Ballon Flight Map
14:58 *Balloon (can't find the link)
14:59 http://www.flickr.com/photos/g[…]99116876/lightbox
14:59 Found it. :)
15:00 Weather was calm with no wind, brought the camera up and down real fast
15:01 Jeff W: Mapknitting can be daunting if you have 30 images. But if you get a balloon real high, you can potentially get your entire area in one photo
15:01 Stewart: It's not easy to fly high, but its great if you can
15:01 1500 feet is great!
15:02 Chris E: Questions about elevations
15:02 Stewart Long: When you're in MapKnitter, its critical to have a base map or a geo reference layer because you have to know where everything goes
15:03 Pat C: How does it differ if you have a mountains or if you have a downtown area?
15:03 Stewart L: It's not different, it's about the base map
15:04 If you think abou flying around taking vertical pictures. They can be union because they're taken from the same perspective.
15:04 Seamless Distortion
15:04 http://publiclaboratory.org/no[…]amless-distortion
15:05 When you have distorted buildings/shapes, you can follow the sidewalks and roadways to match
15:05 Ignore the "stuff" that comes off the ground
15:07 Jeff W: Warp Tool in Photoshop is a powerful tool
15:07 result in a good stitch
15:07 Liz had a bunch of maps in Rio that were on a really steep hill
15:08 Occupy Oakland
15:09 http://publiclaboratory.org/ma[…]d-10am/2011-11-02
15:09 http://publiclaboratory.org/ma[…]d-12pm/2011-11-02
15:09 started in the back of the pack and chased them with the balloon
15:09 Jeff W: as they ran in terror!!
15:10 Stewart: Caught up with the front of the parade. We got the entire parade as we raced through it.
15:10 It was a specialized way of mapping
15:10 everything we wanted to map on the ground was there
15:11 You can see that there are people that are in more than one picture, but we made sure to race through
15:11 Jeff W: Google isn't producing maps with any parades
15:11 Steward Long: Chandeleur Islanda, Louisiana
15:11 http://publiclaboratory.org/es[…]isiana/2010-05-08
15:12 If you know the height, you can rescale it
15:13 Jeff W: discussed adding a feature on Mapknitter that can read information on your photo
15:13 like GPS information
15:13 You can res up by sitting in one place
15:14 If anyone has GPS in images, there can be an initial placement
15:14 Don: Seeing correlation is fairly advanced, where they take points that are common in the picture and correlate them.
15:14 Pictures of Leaning Tower, correlate and you can get a 3D image
15:15 There are packages designed fo rthis
15:15 Pat Coyle: There are services that offer this for a fee
15:15 Citizen Science notion, this is cartography. It's neo cartogrphy. It's people making maps.
15:15 Stewart Long: The human factor is interesting. When you make the subset, it's people deciding what images to make
15:16 In Pasadena, a trolley car went through one of the scenes and the people decided to keep the trolley car in the map.
15:16 You have ultimate control when you do it yourself
15:16 Wilkinson Bay (East), Louisiana
15:16 http://publiclaboratory.org/wi[…]isiana-2010-07-22
15:16 Washington Square Park, New York City
15:17 (project by Oscar)
15:17 http://publiclaboratory.org/ma[…]w-york/2012-10-01
15:17 Used the center fountain and the grass on the outside
15:17 and guess on the middle, because we had no georeference for it
15:17 There's additionally a wild distortion on the arch
15:18 Grand Isla, Louisiana
15:18 http://publiclaboratory.org/ma[…]isiana/2011-02-25
15:18 Using same scale factor to determine the size of the beach
15:19 a single row of images
15:19 Rifle, Colorado
15:19 http://publiclaboratory.org/es[…]lorado/2011-09-16
15:19 two processes that talk to each other
15:19 Jeff W: Zoom in on the animals!
15:20 Stewart: You can see goats and horses
15:20 Jeff W: A great example. This is now the primary image on Google Images. The goats and horses are immortalized.
15:20 Mathew: But the goats and horses may be in a different spot next time.
15:20 Stewart Long: But there are a lot of stones and hard features that will be there through time
15:21 Occupy UC Davis
15:21 http://publiclaboratory.org/no[…]davis-nov-21-2011
15:21 difusion
15:21 zoom in on the people
15:22 100-150 feet above ground level
15:22 resolution is what you can distinguish with each pixel
15:22 with 1 meter resolution you can barely distinguish a car
15:22 with 1 foot resolution you can see a chair
15:23 with 2 centimeter foot resolution you can see people and a bicycle spoke
15:23 You won't really invade on people's privacy because you only see the very top of their head. You can barely tell if its a man or a woman.
15:23 Privacy isn't a problem you run into from vertical imaging
15:23 Today we plan to shoot more outside
15:24 Lumcon Pictures from shots on Thursday
15:24 Open up for Questions on projects you want to map
15:24 Shannon: 10-15 minutes for questions
15:25 Adam: Are people generally figuring out their height by lenght of string? What's the equation for height?
15:25 Stewart Long: If you're at a 1000 feet height, you cover about 1000 feet on the ground
15:25 The focal length of the lens makes it vary too
15:25 Jeff W: a good ballpark is 1000feet high for 1000feet
15:26 Stewart: I measure different things. I measure a feature in the base map and a feature in the new map.
15:26 And do that 4 or 5 times and you know you're average
15:26 MapKnitter will do it for you MAGICALLY
15:26 Jeff W: Looking at the area of the whole area and averaging
15:26 It's not a bad estimate, giving the possible sources of error
15:27 *given
15:27 Stewart: With balloon line I measure 100feet, every 100 feet I put a little tab on my line so I know how far out
15:27 It's nice feedback when making the map
15:27 The more you use your tool kit, the more comfortable you'll be
15:28 Jeff w: Method chosen was ad hoc. It was chosen for part of the function
15:28 There's an advanced option tab in the export window in MapKnitter. You can choose resolution
15:28 There's a histogram of all of your resolutions.
15:29 Adam: Webapp on resolution of photos. Can you explain? The app you drag the files.
15:30 Jeff W: It would be nice to drag and drop into mapknitter.
15:31 mathew I sorted the notes by topic (cities, coastal areas, etc) and added maps from the chat.  check them out here:
15:31 http://publiclaboratory.org/no[…]-talks-map-making
15:32 Hudonnoodles http://archive.publiclaborator[…]tics/dragdrop-js/
15:32 Thanks Mathew!
15:34 Using images from Mathew to demo the drag/drop feature of link above
15:36 Thermal flashlight, the resolution of a given color is the circle of light it shines
15:38 Oops. I just realized that Adam is really Scott in all of the above references. Sorry!
15:39 In the drag/drop app, you can analyze the file size of the images.
15:39 The biggest images are usually the sharpest and have the most detail
15:46 @Liz - how do we end the meeting?
15:47 crabi has quit IRC
15:54 mathew has quit IRC
15:56 liz #endmeeting
15:57 Hudonnoodles #Topic Spectrometry
16:02 Jeff W: PLOTS Spectral Workbench, quick overview and then some quick spectrocity
16:02 http://spectralworkbench.org/
16:03 Showing Video: Public Lab DIY Spectrometry Kickstarter
16:03 http://www.kickstarter.com/pro[…]-spectrometry-kit
16:06 problem statement: we can point at brown stuff and call it oil, but sometimes you want to make a more definitive identification
16:06 taking samples, sending to labs and getting results back
16:07 Scott: We go out in the field to show people the oil, but if we do send it to the Lab it costs $100s to $1000s of dollars to identify our samples are BP Oil
16:08 Jeff: Scientists generally use spectrometry  at some part of their research.
16:08 IF you look at more specific colors, beyond red blue and yellow, there's a real pattern that emerges
16:08 Types of Outcomes: Is this BP Oil or Motor Oil?
16:08 Is this oil or gunk?
16:09 What do we want to prove with a simple tool?
16:10 Jeff: Passing around the newer version of the spectrometer
16:10 To separate out all the colors, you need a prism
16:10 DVD is cheap and almost available anywhere. You can slice a DVD and then glue it to the screen
16:10 the beam needs to hit the webcam and split out all the colors
16:11 The rest is software, post tool
16:11 http://spectralworkbench.org/
16:12 The hard box makes it a bit more sturdy
16:13 Capture online on Spectral Workbench
16:13 You can choose the external, allows
16:13 Flip the spectrometer up at a light, and you can get a reading real time
16:14 @Liz - feel free to add to the notes. I'm not sure if I'm using the most technical of terms. :)
16:16 Typical Output: http://spectralworkbench.org/spectra/show/1153
16:18 http://spectralworkbench.org/spectra/show/913
16:19 Atmosphere of Pfizer, Brooklyn
16:19 We're interested in oil, you can shine a full spectrum light through oil and measure what's absorbed
16:20 But a lot of the good information is off the scale on the infrared
16:20 You can shine a laser through oil (example is a pure green light), but when you shine it through oil sample you see a different color
16:21 *olive oil, extra virgin
16:24 Playing with lasers!
16:25 Focus on Spectrometry Project - measure the color
16:27 eustatic <eustatic!~eustatic@qwebirc.media.mit.edu> has joined #publiclab
16:27 eustatic hey, is this thing on
16:27 Hudonnoodles yup yup!
16:30 A lot of people can build the tool
16:30 now the question is, can we work on the science part of it?
16:31 can we take samples and build credibility with our tests
16:31 eustatic polycyclic aromatic hydrocarbons flouresce in UV.  these are double bonded "benzene" 6-carbon rings.  the "ene" designates a double bond within the molecule.  so "propane" is an "ane", a single bonded molecule.
16:31 Hudonnoodles the Aquarium material is building coral and they need to analyze their lights, so they can utilize the spectrometer
16:31 hopefully that can build a momentum to some of the harder cases
16:32 eustatic so propane would not flouresce? that is not clear to me
16:32 Hudonnoodles Flame Emission Spectrometry - heat things really hot and it will glow so you basically only see atoms
16:32 eustatic but propane, butane, etc are not PAHs, because they are not "cyclic"
16:35 Hudonnoodles Jeff W: we can do what we want. we can build, we can do tests, we can boot up spectralworkbench
16:35 Its up to the group, what should we do? Or break up into group and try them all.
16:35 Film strips, with infrared photography
16:36 Dan Beavers: Lets play with stuff!
16:36 Group is now playing with stuff.
17:37 shaief <shaief!~shaief@qwebirc.media.mit.edu> has joined #publiclab
18:22 shaief has quit IRC
18:23 Hudonnoodles has quit IRC
18:23 Hudonnoodles <Hudonnoodles!~Hudonnood@qwebirc.media.mit.edu> has joined #publiclab
18:23 Hudonnoodles #topic The State of Kite & Balloons for Mapping by Mathew L
18:25 Japanese Kite Festival Image
18:25 What would it be like to have an integrated culture of kites for mapping?
18:26 In Japanese Image, mostly young people holding a handmade kite
18:26 The kites are rectangular, easy to make
18:26 Most Japanese kites are created with only one knot
18:26 Giant flying kites
18:27 Flying is fun, which is what we should always include in our kite flying
18:27 It's a social activity
18:27 engage with community while gathering images
18:27 liz has quit IRC
18:27 Hudonnoodles Image on Screen: tyvek kite by splitting bamboon
18:28 Currently, we're working on kites that are easier to build
18:28 "You can tell that it works because people are happy." -Mathew
18:29 Observational Balloons
18:29 meetar <meetar!~meetar@qwebirc.media.mit.edu> has joined #publiclab
18:30 Hudonnoodles Balloons and Kites are possibly the least aggressive technologies you can have
18:31 Image on Screen: One of Stewart's balloons with four attachment points
18:32 "Wouldn't it kick butt if we could fit the entire kit into a briefcase?" -Mathew
18:33 #topic Balloon & Kite Mapping
18:33 Field Trip, Entire team is out mapping Lumcon
18:33 Estimated time of return = 4PM
18:35 eustatic has quit IRC
18:42 liz <liz!~liz@qwebirc.media.mit.edu> has joined #publiclab
18:50 liz has quit IRC
20:27 Hudonnoodles #topic Spectrometry Continued
20:28 Jeff Warren and a group are testing the spectrometry of bunsen burner in Room 202
20:28 *of a
20:46 warren has quit IRC
20:53 Hudonnoodles_ <Hudonnoodles_!~Hudonnood@qwebirc.media.mit.edu> has joined #publiclab
20:53 Hudonnoodles_ #topic Indoor Air quality and Hydrogen Sulfide Detection
20:54 Dan Beavers: pre-discussion talk. One Laptop per child.
20:54 Founder wanted to create a laptop that cost less than $100 per computer. HE created a laptop that has wifi and meshing.
20:54 It has a sunlight readable screen that used reflective technology
20:55 Inside the screen is colored, and outdoors it's black/white
20:55 It uses sugar as an operating system. Dual boot.
20:55 http://one.laptop.org/
20:56 Hudonnoodles has quit IRC
20:57 Hudonnoodles_ has quit IRC
20:57 Hudonnoodles <Hudonnoodles!~Hudonnood@qwebirc.media.mit.edu> has joined #publiclab
20:59 Hudonnoodles #topic Air Quality Monitoring & Hydrogen Sulfide Testing
21:00 Shannon: Black and white photography paper. Covering it with a solution and exposing it to hydrogen sulfide
21:00 pre-test in New Nexico
21:01 Preliminary testing right now, to see if it's possible to do
21:01 Don and JEff have started to talk about developing a sensing component to do some calibration for this test
21:01 The New Mexico group is slightly stalled
21:01 Trying to combine the communities working on this project
21:02 Jeff W: Sara ordered hydrogen sulfide sensing components. Just the sensor, no electronics.
21:02 For a long time, it seemed like no one knew how to connect to them.
21:02 They're silver and $70/piece
21:03 http://publiclaboratory.org/to[…]n-sulfide-sensing
21:03 It's sensitive to humidity and temperature
21:03 To make our own, we would have to also buy a humidity and temp sensor
21:04 More info is on the link above (http://publiclaboratory.org/to[…]-sulfide-sensing)
21:04 No matter how much work we do, we may not get a super sensitive sensor
21:05 Dan Beavers: How much do the dreagers (sp?) cost?
21:05 Shannon: 10 for $90
21:06 In our first test we used a lot of them because we would tape them to PCV tubes. The hydrogen sulfide in the swamp wouldn't be high enough to be picked up by the tubes.
21:06 *draeger tubs
21:06 *draeger tubes
21:06 We need to figure out the industrial side of things.
21:07 No change in color on the tubes, even when we left out for 8+ hours
21:07 They're not the pump draeger tube, all you need to do is crack the tubes off.
21:08 Don: There seems to be a saturation issue, if you leave it out for 2 days does it max out?
21:08 Shannon: With the draeger tubes, it maxes out at 8 hours
21:10 Dan Beavers: When you take the cap off, you expose it to light? Right?
21:10 Shannon: Apparently it only is exposed when it comes in contact with...one specific element.
21:10 Need to go back to basic questions.
21:14 We need to further the discussion with more people
21:14 Dan Beavers: Is it light sensitive?
21:14 Shannon: Potentially
21:15 Looking at visual on screen
21:15 Jeff W: There should be a correlation between distance and darkness
21:15 Dan: and concentration
21:16 Another factor is wind dirctor
21:16 osbock <osbock!~osbock@qwebirc.media.mit.edu> has joined #publiclab
21:16 Hudonnoodles Jeff W: It may be better to look at it as a heat map
21:16 Don: At UMASS we were discussing about calibration and how it could be very toxic
21:16 Dan Beavers: Isn't it just a natural gas?
21:17 Shannon: Yeah, it is. But it also can be dangerous.
21:17 osbock has left #publiclab
21:17 Hudonnoodles Jeff W: Human health hazard is 5+
21:17 Jack Summers, researcher
21:18 http://publiclaboratory.org/people/j-summers
21:20 Trying to find the cost of a glove bag
21:20 Jeff W: At this point, we're testing the tool and not the site
21:20 Shannon: There's two issues. The network of people working on it is not broad enough yet.
21:20 If we go to Louisiana, it would be 5 hours from New Orleans.
21:21 MExico City doesn't want to move forward until they have standardized safety procedures in place
21:21 Jeff W: The sites that they were looking at, they may be very sure that there's hydrogen sulfide. In some ways you don't want to do anything that there's uncertainty
21:21 At this point, it seems like we should do it in a closed chamber
21:21 with some rotting compost
21:22 Jeff W: With spectrometry, I've been doing a lot of testing with Olive Oil. Although no one is interested in olive oil. It's still a super safe material and kid friendly.
21:24 http://www.sigmaaldrich.com/la[…]ich-atmosbag.html
21:24 Oscar: How serious of a problem is it?
21:24 Shannon: with the boom with hydro fracking, its becoming a bigger problem with air emissions
21:25 If you're severely exposed to it, you can be killed on the spot. It's primarily a problem for employees at oil sites.
21:26 Alexandra: You can smell it at very small concentrations. But if you get enough of it, you can no longer smell at all. And at that point, it's almost too late because you can die soon after.
21:26 Sensors are $100 to $600
21:27 We want to detect.
21:28 Alexandra: http://en.wikipedia.org/wiki/Hydrogen_sulfide
21:28 Jeff: What is the dream and awesome sensor we want to create?
21:29 Throw it on the ground and come back to itt the next day? Yes/No dangerous level?
21:29 Or do we want to throw it on a string?
21:29 Let's think about the use before the technology.
21:29 AND. What is the "Hello World" for this?
21:29 Just to show that the tool is working
21:30 Don Blair: With this project, it may be a combination. IT may be a piece of paper with a binary type result
21:30 Then to follow up with a more sophisticated sensor if you have a bigger issue
21:31 Air Glove Bag: http://pubs.acs.org/doi/abs/10.1021/ed072p96
21:32 Jeff W: http://publiclaboratory.org/to[…]r-quality-mapping
21:32 A very beginning proof of concept
21:33 Whatever the case, if we can find sensors that will detect this, then we want to try to find a way to cheaply produce maps of your home to show the sources
21:33 This is related to the thermal flashlight project
21:33 The idea is to put a gas sensor and have it shine a color light
21:33 https://www.google.com/search?[…]&biw=1411&bih=710
21:33 Example: photographic genre
21:34 They put a light on top of their roomba
21:34 There's no data except for the movement of the roomba
21:34 We thought, could we make teh color of the light change with the presence of some chemical
21:34 The thermal flashlight is this, but with heat and without the roomba
21:34 The work of the students at RISD, they were thinking about gas leaks in apartment.
21:34 They bought a few sensors and assembled it
21:35 The sensor isn't so specific
21:35 It changes on the scale of feet, not inches
21:36 The roomba is good at covering space
21:36 It would be interesting to put the sensor in the throat of the roomba
21:37 Another idea is to put into a hamster ball with legos
21:37 It didn't get stuck on things
21:37 It detected legos, but it wasnt very reliable
21:39 formaldehyde sensing - we need to think about a handheld or wand sensor because it's more than ground level sensing
21:42 Commercial Glove Bags: $30-$40
21:42 liz <liz!~liz@qwebirc.media.mit.edu> has joined #publiclab
21:43 Hudonnoodles http://spot.us/pitches/1535-tr[…]al-housing-crisis
21:43 Looking at FEMA trailers post-Katrina
21:44 Jeff W: We should talk to more people working on similar projects
21:45 For example, people who live near a smokestack with burn off. Its not something Jeff W. sees everyday.
21:46 Shannon: part of the idea is to start the idea with formaldehyde. With mold, do you need a filter?
21:46 Mathew: Cleaned out his vents and found mold. They were sick all winter from it. May be of interest.
21:47 And Liz Barry found mold in housing complex. Only labs can test for mold. But it made Liz challenge the assertion.
21:48 @Liz we were just talking about you! :)
21:49 Jeff W: How do people test for asbestus?
21:49 Dan: There's a filter.
21:50 Don: What are the dangers? And what are the tools for monitoring?
21:50 A localized guide.
21:51 Jeff W: A really good problem statement is the first step.
21:51 Maybe it would be really useful to come up with a set of questions to help frame the problem?
21:51 Dan Beavers: a flow chart
21:51 Jeff W: For example, is this dangerous to human health?
21:52 OSHA Standards: http://www.osha.gov/law-regs.html
21:52 Are they enough?
21:52 What are the questions you want to ask to frame a problem? (when developing a tool)
21:52 --Is it dangerous?
21:53 --In what quantities?
21:53 --Where is it commonly found?
21:53 --Is it air? Or liquid? Or soil? Or solid?
21:53 --In what concentration?
21:54 --How is it currently being tested? How much does it cost?
21:54 --Who comes in contact with it?
21:54 --How people come in contact with it?
21:54 --Exposure time?
21:54 --How prevalent is it?
21:54 --Localized? What geographic area?
21:55 --Emergency Response Guide Book, FEMA (all chemicals that are transported and how to deal with hazards)
21:56 --What are the resources to answer some of these questions?
21:58 http://www.tc.gc.ca/eng/canute[…]uide-menu-227.htm
21:58 The entire booklet above
22:06 liz has quit IRC
22:07 mathew <mathew!~mathew@qwebirc.media.mit.edu> has joined #publiclab
22:09 Hudonnoodles_ <Hudonnoodles_!~Hudonnood@qwebirc.media.mit.edu> has joined #publiclab
22:10 Hudonnoodles_ #Topic Review of Photogrammetry Class
22:11 On Screen: National Wetlands Inventory map from the U.S. Fish and WIldlife Service
22:11 Different plants are going to have different texture
22:13 Chris: Is there ever a resolution of finding out what all that green actually is?
22:13 Scott: Shrub, Fresh watered forested shrub...
22:13 Some of the green is just marsh, then there's shrubs
22:14 Hudonnoodles has quit IRC
22:17 Hudonnoodles_ Signature Development
22:17 Different spectra to draw lines
22:17 What is the more human way to do it?
22:17 Handout: Signature Development Assessment Summary
22:18 Includes Photo Characteristics, Explain, If yes - characterize
22:18 Mathew: Is there an image reference for each of these?
22:18 Jeff W: Or a MapMill feature
22:18 Maybe you would be asked, "Choose a shape for this feature?" OR "Choose a color for this feature?"
22:20 Dan: Metadata needs to be discussed with the images
22:20 What do you want to do with it?
22:20 Scott: Metadata is necessary, but no one wants to do it
22:22 Similar color/texture
22:33 Color indicates how much photosynthesis is going on
22:34 Depending on the season of the flight, you're going to get different colors
22:47 mathew has quit IRC
22:49 Hudonnoodles_ Scott: There is an issue with metadata and needing to do some of it manually
22:49 Any other questions about interpreting aerial imagery? How to find out, what is it?
22:53 Clashifier
22:53 clashifier.publiclaboratory.org
22:54 It doesn't work very well in this version, but it was an try
22:54 It's a way to classify the image
22:54 It's only looking at red, green, and blue
22:54 Based on someone making that model, you can classify
22:54 IT colors the whole image based on the training
22:55 It doesn't work significantly well, but it's a work in progress
22:56 It's a crude classification
22:58 People are much better than computers at classifying and identifying plants
23:10 Hudonnoodles_ has quit IRC
23:17 Hudonnoodles <Hudonnoodles!~Hudonnood@qwebirc.media.mit.edu> has joined #publiclab
23:18 Hudonnoodles #topic Happy Hour
23:18 #endmeeting
23:26 Hudonnoodles has quit IRC

 « Previous day | Index | Today | Next day »     Channels | Search | Join

Powered by ilbot/Modified.
Webmaster