My Progress
I took another round of demonstration images of trash around the neighborhood using the new system. Mixed results. The series of images shows a straw in medium vegetation, then a picture without the straw, then the straw on a measuring board to get the rough dimensions, then the straw on a green-screen like background.
I used a tripod of sorts, but I couldn't get the overhead shot I wanted with it without getting my glove and/or shoe in the way. I think the solution is going to be to shoot at a 45 degree angle so I can use the tripod as it was designed. I don't have the money/time to design what I think I'd need for an overhead shot so I'll make do for now. I was hoping to avoid an angle shot because of scene distortion at that angle. Maybe it's better to have the angle anyway, since it's a more realistic representation of how a scanning camera system might look for trash.
I took 21 photo sets of various trash and documented it in a spreadsheet. That part seems fine (for now). Taking some of the free neural network / AI courses online has helped me understand what good practices look like for this particular challenge.
Next week I'll try revision 3 of the system and see if things work out a little smoother.
Ocean Voyages Institute
I came across a new article that this institute (that I had not found through searching) was featured in. They recently came back from a voyage out to the Pacific Garbage Patch where they collected 40 tons of plastic, including a 5-ton ghost net. The picture is impressive.The article mentioned something I had been considering: they asked people traversing the patch to attach GPS trackers/broadcasters to nets they found. These could be tracked over time to watch their journey and, when sufficient densities of nets justified the expense, they could go collect them. You could also clean up anything you found along the way. As part of their campaign, they also used drones to do route mapping and surveillance to find things.
Unfortunately, this journey cost $300,000 and took 25 days. (They sailed most of the time so their top speed was reduced.) $300K for 40 tons is $7,500 per ton of trash. The estimate is that 8 million tons of plastic enter the ocean each year, so this would be $60 Billion dollars per year required using this method. Hmm... Somehow this number needs to fall by a factor of (roughly) 7,500.
I think it's pretty cool, though, what they're up to. I'm glad they made headlines and that I can add them to organizations to watch. I'm hoping that they find a way to get their costs way down so they can keep doing their work and finding efficiencies.
https://www.oceanvoyagesinstitute.org/
Comments