2019 Present and Future? Campaigns


#1

Is anyone working on “Western Fires”? Just wondering if it is being completed? and once it is…what is going to happen next. I have found it tough to stick with that campaign and cannot load H. Maria.
We have not heard from @Mel_Nod in a long time…wondering what is going to happen next with TN?
Any ideas?


#2

Sigh. I said I wrote to them. ;-( I’ll let everyone know when / if I hear back.


#3

I was checking in with everyone what they thought…(I saw your note posted on the other topic).


#4

I have been working on the Maria Campaign. It loads quickly when I use one of my alternate accounts, I did this by creating a new username with a different email address (which I created in gmail for free. Maybe Tomnodders will need to do that to get Maria to load).

I am happy to loan my kateg*s that haven’t been used in this campaign by changing to a dummy password that you could then change to a password of your choosing.Let me know if you want to do this. Then you wouldn’t have to create a new account and you could give it back by changing the password again and handing off to me. (Just realized, I would have to create a dummy password for the gmail account as well so you can access it, but that is fine with me if this helps finish the campaign. By the way, when you create a gmail account, you don’t have to provide your correct info - i use shortened version of my name And a birthday slightly off. The only time it required a phone to text a confirmation code was when I tried to create more than one new account in a a row. My intention is to eventually delete them, I don’t use them for anything else other than to tie to a new Tomnod account


#5

I had a little bit of free time after Christmas… I had fooled myself to thinking I was a bit further on, but no…

Anyway I went back to using the ‘rules’ I defined as how I read them… based on what I was seeing as well.
I’m still playing that way… so I’m mostly consistent… most of the decision making is in the Colorado fire zone.
Have been known to waffle there a lot…

Anyway, no great effort has been possible due to eyes and time available… I should be approaching my 10% of the tiles low limit… and will keep at it beyond that… I have to re check what @Jim7 said it was… total
exactly… somewhere in the low side of 33 to 35 k polys I believe off the top of my head…
Probably some sort of testing going on… or just something to do… so I play at that way and added the bit of locating the fires and plotting the outside of the newer images… and in a few of them checking out to ID the times of the before images… sometimes very recent… and one was years prior and poorly mated to the newer image in terms of registration… I spent some time playing with defining an offset to use, but it seemed to change a fair bit… the CA fires were much better matched. CO was a fair bit of work due to the changes of several years not to mention capture quality and angles… but that is the nature of even conventional air photography and cameras in general.

One can always find something to self educate on… I keep going and reviewing my textbooks to see if I can Identify a base method being used… or trying to find out what is going on if I can’t see one there…
Search theory can be quite simple or complex and fun to find better ways to get the best coverage for the manpower resources. Or even for one person to do their coverage… we see that some like one method and some like another… but there is a base that says pick a pattern and do it… or even change it once in a while to help get it all quickly… all that means is if you work left to right in a grid, you are in fact doing the same thing as the guy who does it right to left, up to down, down to up or even get spacey doing a spiral… it is the same…
even if you swap them for a change (which can be good for morale too)…

Point is… there is always a better way to some people… yes computers can scan things… faster… but are they as good… who knows… WE should seek ways to make sure things happen quickly and accurately in whatever we do… even if it is teaching the computers a bit more… but they won’t be at the level of human insight for quite a while yet. And they do have the limits of being able to move around, see some things clearly enough to work from… but I can assure you that they can do much better than we get to work with…
Our strength is working the blind spots where possible… even if it is just identifying where those blind spots exist is good… I believe in the recording of nil data… it’s as good to know where something is NOT as much as where it IS at certain points of the search… especially the very onset of the search and each sub stage as well… That is part of how I locate the images… at first I know nothing about them, and I keep a record of what I see as I start to browse around… marking objects or judging a poly… when I need a break I start trying to see what I have… how many image strips for one… in this case I have 3… and not seen any evidence of others… two have really easy to find landmarks and got done quickly… based on what i could see in them. and a bit of educated guessing… the CO map had some things it took time to ID… for one thing I had quickly ruled out being CA… but the west is big. using new and older images, I soon saw that the old covered a much larger area and saw things that eventually allowed me to zero on… but I only worked on the new image strip… since the other image strip is simply around the newer one…thus fairly easy to add later on.

Like I said NOT vs IS is good.

Anyway, always something happening… and I have identified some new personal limits as well…


#6

Good for you @Doug4. I have had trouble sticking with it. I will keep trying to add a few more here and there. I’m going to try another way to get into H. Maria…hope it works. Otherwise…hoping for something new … eventually.


#7

I gave up as the pink area is usually so tiny that one cannot reasonably determine ‘burned’ or ‘not burned’, only from the surrounding area might one guess. The area is so small that sometimes there is nothing except a pink blip. I have added a screen shot…

http://www.tomnod.com/campaign/campaign_2390/question/9784?ref=6994120


#8

Hi Rosemary, I know some of the polys are very small while others are not. That is caused by the machine learning algorithm (AI). If you use the “+” at the upper right side you can zoom in. Usually just +2 is all that is needed to make a proper determination. Sometimes just where the poly is - completely surrounded by scorched earth - you can deduce that the area within the poly is burned. If you see green in the area around the poly, then you will probably have to zoom in. The extra step of zooming in takes a little time - I know because I’ve completed this entire campaign. :grinning: Hope this helps.


#9

I wonder why we have had campaigns for the Bangladesh camps, the African camps, etc, however we have not had any campaigns for the migrant camps in Greece, Spain…

https://www.unhcr.org/greece.html


#10

An agency - government or private (charitable) - would have to contact Digital Globe and request it. There’s so many times we wish we could do something, but unless there’s a request (as stated above) made, we can only sit and wait. Frustrating at best. :neutral_face:


#11

Because we did UN tents/huts in the other campaigns and they are present in these camps also, it would seem it would help the consistency of their administration in all areas of the world.

As technology, business and market trends are changing quickly, maybe a change in this method would ensure their future in the marketplace.


#12

True, but as you can see by the article you posted, Spain is just one more country that increased the number of people in a camp by over 500, thus the camp grows. This growth can be on anywhere from a daily basis to a weekly basis. It would be hard to tag huts/tents when more and more are erected constantly. It’s hard but could be done. All a group would have to do is approach DG and ask. The waiting is indeed frustrating.

Aside from this, while working on H. Maria, I’ve noticed where there were different shades in the vegetation in the “before” maps that ended up being flooded or had landslides in the “after” maps. This could also be a data set that planners could use to limit development in certain areas. In other words, these areas can be considered as flood prone or flood plains. It was terrible to see houses on one map and nothing but a totally wiped out/washed out area in the other.


#13

True, however Bangladesh was continually changing also. Greece and Spain would be a much smaller scale and easier to identify the needs.

As in H. Marie, I believe we are doing the same on the Western Fires. It is not so much about the fires themselves, as the fuel (vegetation or structures) and path of the fire for future restoration.


#14

There are also some suggestions from others under the 2018 heading .

2018 - need info about Western Wildfires: Post-Fire Support

I agree in some cases, “your guess” is the only thing possible.


#15

Thanks Jim, everyone,
Yes I realise the zoom helps, unfortunately it is inconvenient and totally time consuming to zoom in every single time. It appeared illogical to have minuscule areas to identify when a whole hillside is burned. It was easier to wait for the next project. Maybe if I apply some self-discipline then the project would get completed and a new project might appear :slight_smile:


split this topic #16

A post was merged into an existing topic: 2019 Jan - WorldView 4 Satellite lost “axis of stability” - 2 yr old can’t make pictures!


#17

Like the recent plane crash with the Argentinian Soccer player aboard, we have had several requests via the forum for immediate help. I hope in the future, through some kind of technology, it is possible to do this.