jlmcgraw
Filing Flight Plan
- Joined
- Apr 18, 2014
- Messages
- 5
- Display Name
Display name:
Jesse McGraw
All,
If you've ever used geo-referenced instrument approach plates in any of the modern EFB programs you know that they can be a great aid to situational awareness. Unfortunately they're usually a relatively expensive add-on, often costing as much as the base program itself.
In the hopes of fostering some cheaper alternatives, and as a spur to all of the programmers out there, here is a link to a spreadsheet containing geo-referencing information for approximately 11,000 FAA/Aeronav instrument approach plates from the most recent 56-day cycle (1404)
https://docs.google.com/spreadsheets/d/1jnHNV3jQWQDGd0d_96k3yNMrxTQf2EnvECVyTYCdtRg/edit?usp=sharing
This data was generated using an open source perl program I wrote which is hosted on Github (https://github.com/jlmcgraw/GeoReferencePlates)
My intent is to make this information freely available in order to increase general flying safety, so please do NOT charge for it in any way should you choose to redistribute it or use it in a program. Please provide some attribution for the source of the data as well.
Some notes about the data:
- It isn't perfect and comes with no warranties about accuracy whatsoever. Don't bet your health or your property on this data!
- That being said, it is pretty good. Here are some numbers from a recent run:
18358 Total plates (doesn't include minimums etc)
- 5975 Plates that aren't to scale or currently in scope (SIDs/STARs, LAHSO, HOTSPOT, Airport Diagrams)
-------
12383 Potentially possible to georeference
- 1093 Military plates (these have no actual text in them and are rendered differently so they don't work with my program)
-------
11290 Currently possible with this program (basically all Civilian charts using any type of approach)
- 161 That simply don't georeference (don't pass basic sanity checks due to mismatched ground control points or the program was currently unable to find enough ground control points)
------
11129 Plates that get georeferenced (a small percentage of these are definitely inaccurate even though they pass the basic sanity checks I've devised so far)
- Note that the PDFs have been renamed to a STATE-AIRPORT-PROCEDURE format from their original names. The program depends on this format, hopefully it won't confuse anyone too badly
- Cells highlighted in red indicate that the process definitely didn't work for this particular plate (no actual text in PDF, bad longitude/latitude ratio, not enough valid ground control points)
- The PNGs I've been using were rasterized from the PDFs at 300 DPI. If you use some other DPI then you'll need to adjust the X and Y pixel scales accordingly. The longitude/latitude extents of the raster won't change
Other tidbits
- The military plates are sourced from a different environment than the civilian ones even though they're both downloadable from the same site. This isn't particularly useful information but it is interesting.
- The published obstacle database heights are not what is always depicted in these plates, perhaps they're occasionally rounded up via some formula. I haven't seen a case yet where an obstacle on a plate is marked as being lower than one in the published obstacle file but it wouldn't surprise me if there was one.
I've only tested this with Ubuntu Linux, so it is probably easiest to use it or one of its variants (xubuntu, lubuntu etc.) in a virtual machine. Virtualbox works great for this.
First steps to get this running in your Linux environment (no testing was done with Mac or Windows, please feel free to contribute patches)
- Install git (sudo apt-get install git)
- Clone the repository (git clone https://github.com/jlmcgraw/GeoReferencePlates.git)
- Follow the rest of the instructions in the readme
I think a good next step would be to crowd-source verifying the accuracy of these georeferences and creating ones for those plates that this program didn't work with. This would probably be reasonably straightforward via a web site or an add-in to existing EFB programs on Android or IOS that have a large existing user base but it isn't something I can work on currently.
If you'd like to see how these look on a map and are familiar with the free GIS program called QGIS, check out my aviation map project here: https://github.com/jlmcgraw/aviationMap.git
Load the example Virtual Raster files (*.vrt) to see how the plates match up with the airspace, runway, obstacle, fix, and navaid data from the FAA.
If you have any ideas about how to make this program or data better please let me know via email or contribute patches at github. It is still a work in progress and I am definitely not a professional programmer so I'll take all the input I can get.
I look forward to your feedback and fly safe!
-Jesse McGraw
jlmcgraw@gmail.com
If you've ever used geo-referenced instrument approach plates in any of the modern EFB programs you know that they can be a great aid to situational awareness. Unfortunately they're usually a relatively expensive add-on, often costing as much as the base program itself.
In the hopes of fostering some cheaper alternatives, and as a spur to all of the programmers out there, here is a link to a spreadsheet containing geo-referencing information for approximately 11,000 FAA/Aeronav instrument approach plates from the most recent 56-day cycle (1404)
https://docs.google.com/spreadsheets/d/1jnHNV3jQWQDGd0d_96k3yNMrxTQf2EnvECVyTYCdtRg/edit?usp=sharing
This data was generated using an open source perl program I wrote which is hosted on Github (https://github.com/jlmcgraw/GeoReferencePlates)
My intent is to make this information freely available in order to increase general flying safety, so please do NOT charge for it in any way should you choose to redistribute it or use it in a program. Please provide some attribution for the source of the data as well.
Some notes about the data:
- It isn't perfect and comes with no warranties about accuracy whatsoever. Don't bet your health or your property on this data!
- That being said, it is pretty good. Here are some numbers from a recent run:
18358 Total plates (doesn't include minimums etc)
- 5975 Plates that aren't to scale or currently in scope (SIDs/STARs, LAHSO, HOTSPOT, Airport Diagrams)
-------
12383 Potentially possible to georeference
- 1093 Military plates (these have no actual text in them and are rendered differently so they don't work with my program)
-------
11290 Currently possible with this program (basically all Civilian charts using any type of approach)
- 161 That simply don't georeference (don't pass basic sanity checks due to mismatched ground control points or the program was currently unable to find enough ground control points)
------
11129 Plates that get georeferenced (a small percentage of these are definitely inaccurate even though they pass the basic sanity checks I've devised so far)
- Note that the PDFs have been renamed to a STATE-AIRPORT-PROCEDURE format from their original names. The program depends on this format, hopefully it won't confuse anyone too badly
- Cells highlighted in red indicate that the process definitely didn't work for this particular plate (no actual text in PDF, bad longitude/latitude ratio, not enough valid ground control points)
- The PNGs I've been using were rasterized from the PDFs at 300 DPI. If you use some other DPI then you'll need to adjust the X and Y pixel scales accordingly. The longitude/latitude extents of the raster won't change
Other tidbits
- The military plates are sourced from a different environment than the civilian ones even though they're both downloadable from the same site. This isn't particularly useful information but it is interesting.
- The published obstacle database heights are not what is always depicted in these plates, perhaps they're occasionally rounded up via some formula. I haven't seen a case yet where an obstacle on a plate is marked as being lower than one in the published obstacle file but it wouldn't surprise me if there was one.
I've only tested this with Ubuntu Linux, so it is probably easiest to use it or one of its variants (xubuntu, lubuntu etc.) in a virtual machine. Virtualbox works great for this.
First steps to get this running in your Linux environment (no testing was done with Mac or Windows, please feel free to contribute patches)
- Install git (sudo apt-get install git)
- Clone the repository (git clone https://github.com/jlmcgraw/GeoReferencePlates.git)
- Follow the rest of the instructions in the readme
I think a good next step would be to crowd-source verifying the accuracy of these georeferences and creating ones for those plates that this program didn't work with. This would probably be reasonably straightforward via a web site or an add-in to existing EFB programs on Android or IOS that have a large existing user base but it isn't something I can work on currently.
If you'd like to see how these look on a map and are familiar with the free GIS program called QGIS, check out my aviation map project here: https://github.com/jlmcgraw/aviationMap.git
Load the example Virtual Raster files (*.vrt) to see how the plates match up with the airspace, runway, obstacle, fix, and navaid data from the FAA.
If you have any ideas about how to make this program or data better please let me know via email or contribute patches at github. It is still a work in progress and I am definitely not a professional programmer so I'll take all the input I can get.
I look forward to your feedback and fly safe!
-Jesse McGraw
jlmcgraw@gmail.com
