Accuracy of Ariel Photography? 1

[Col]

What is the best accuracy available with ariel photography of rural areas and how does it compare with traditional surveying techniques? Can ariel and traditional be readily merged?

 

[carlorom]

You don't stipulate what type of aerial photos you want to use. This is a vast subject known as Photogrammetry. Accuracy depends on height at which photo was taken, type of camera and lens, whether camera was pointing directly down or obliquely, variation in terrain elevation, accuracy of control points on the ground etc.

In a nutshell though one can get mapping done to around 0.1m with proper techniques. There is a technique which converts aerial photos to a proper map called Orthophotos.

We can obtain 1/10000 scale orthophotos which will allow scaling directly off the map to around 1m accuracy.

Generally it will be more accurate to use traditional ground survey methods as long as the site is not too large. If however vast tracks of land need to be mapped it will probably be more efficient to use aerial mapping techniques.

Lately satellite remote sensing data is also available and approaching accuracies mentioned earlier.

You could probably obtain specific data of accuracies available in your area by contacting your government mapping agency or your local Land Surveyor, or photogrammetry association.

 

[Col]

For orthophotos with a ground model and an stated accuracy of +/-1500mm, are the errors in the ground model normally distributed? Has there been any research on this? If so what spreads are usual?

 

[sc]

May be a bit late but ....

Our company has been using for the past 2 years a high accuracy GPS system (+- 10mm) for rural applications. This system is vehicle mounted and can be used at a travel speed of up to 25km/hr with reasonable accuracy (+- 25mm) which is generally better than required by the client. This accuracy is for both horizontal and vertical position.

This is far more cost effective than aerial photography due to time savings and cheaper equipment utilised.

One of our client has obtained the photographs and the GPS obtained contours and combined the 2 in CAD with a reasonable overlay result.

sc

 

[carlorom]

SC

Sounds quite interesting what you are doing but I would be more interested in knowing which GPS you are using and in what mode - Heights on a moving vehicle - how accurate are these if vehicle bounces? Are you using two receivers (one being stationary) Your results may possibly be adequate for the type of work you do but I doubt if you can be sure of getting 25mm accuracy - yes maybe between successive points but how well do those compare with other points say 2km apart - Before I say any further please elaborate a bit more on how you use the GPS receivers and data processing

 

[IRstuff]

Differential GPS is the usual approach for such high accuracies. This is the accepted method for bypassing the deterioration of the accuracy that's built into GPS design. You provide a reference receiver that sees the same GPS constellation that the mobile unit sees. The reference receiver is accurately surveyed for absolute measurements, but in any case, will send out corrections to the mobile unit.

Depending on the terrain and visibility of the GPS constellation, 2km spans should be quite doable.

TTFN

 

[carlorom]

Yes that is theoretically correct and accepted practice I however query the accuracy obtained physically from the actual position of the vihicle on the terrain - if it bounces around you may well get a spotheight of a point above the ground for instance. Also how frequently are spotshots recorded when moving vehicle - one could thus "miss" a point due to the speed of vehicle.

The point I am trying to make is that the survey methodology must be designed to befit the accuracy for which the job is necessary. In general I thus agree with what you are doing but be aware of the pitfalls.

Regards

 

[flamby]

Though the accuaracy of photogrammetry is becoming increasingly reliable, and I have seen some road project authorities using it for preliminary surveys and estimation, the time is still not ripe for use of areal surveys for detailed design and engineering. So the answer is no, you can marry these different pieces of data and use it with reasonable accuracy for engineering. The conventional survey instruements give to accuracies to the fraction of mm whereas the BEST you can hope to achieve from areal surveys done with best of the technology will be 25-50 mm. Hence the line is clearly drawn. Use Areal data for project formulation, planning and preliminary costing. But that done, you have to go out to the field with your old trustworthy instrument. As on date, there is no alternative to tripods.

 

[IRstuff]

Plus, aerial surveys cannot reasonably give you the lay of the land, e.g., slope, etc., which can only be accurately gotten on the ground.

TTFN

 

[Bayou]

DOQQ aerials are now used by USGS for mapping purposes.

http://mapping.usgs.gov/internal/af/Admin/Policy/92nmd10.html

They used these typically 1m pixel resolution aerials to create their digital line maps. Therefore the aerials have to be within 33.3 feet on a 1:12000 scaled map, or 1/30" on a paper map.
More accurate aerials can be produced based on a higher resolution. Houston has 0.5m pixel aerials set to the same DOQQ process as USGS. Therefore they are 4 times as accurate with the higher resolution. Plus when the aerials were flown inassociation with a LIDAR run. Right now LIDAR is taking about 40000 data points per second.