Ordering Info
Before any images can be acquired, you must make some specific decisions to make sure that you get the images you need at the lowest possible cost. First you must decide what resolution is required. Then you must decide what accuracy you need. Finally, you must decide how you will obtain the information you must provide to enable us to acquire your images.
Cost/time tradeoffs
Before you can use an image as you intend, you must ensure that the resolution of the image will allow you to see what you are looking for. Resolution defines the smallest object that can be seen on an image. Naturally, a smaller pixel size will provide more detail and therefore higher resolution. You can often get by with an image that has lower resolution, but the time to interpret increases.
Whenever you look at an image, it takes some amount of time to interpret what you see. For example, you may want to determine whether a rectangle on an image is a parking lot or an open field. With high enough resolution, you can easily see the parking lines and end up spending very little time determining that the rectangle is a parking lot (see image below). On the other hand, consider looking at an image that does not have enough resolution to see the parking lines. You can look elsewhere on the image to see how the roads and building relate to the area and eventually determine that it is a parking lot. But this scanning around for clues related to the area in question takes additional time before you can determine what it is, and furthermore is subject to misinterpretation.
This is the time to interpret. An image that has lower resolution may still provide the information that you require, but at the cost of extra time to interpret and potential interpretation errors. Eventually, the resolution can become so low that no amount of time will allow you to interpret the image.
How much re solution do you need? That is up to you. There are some things you can do to determine the resolution you require. You can ask for sample images from a potential image vendor. We have many high-resolution images to help you out. In the Samples page, you will also find some images that can help you determine your resolution needs.
Most of our customers use 30-cm to 60-cm resolution (or 1- to-2 foot resolution). As with accuracy, the higher the resolution (smaller the area covered by a single pixel) the higher the cost. The reason the cost increases is that the camera has a fixed number of pixels in its sensor. To increase the resolution, the amount of ground captured in each image decreases. This can require more images to capture the same subject.
At times, the size of the subject is small enough that an increase in resolution does not increase cost. A larger subject may require images to be combined together to cover the entire subject, increasing cost via additional processing.
Cost/accuracy tradeoffs
Accuracy costs. It is easy to specify a very high degree of accuracy to make sure that you have the accuracy you need. But this can vastly increase the cost of the image. The highest degree of accuracy requires that you have ground control points for each image. A ground control point is a survey marker that can be seen in the image and is placed before the images are taken. This is very expensive as ground crew must travel to the site, locate the ground control points, and find or create something that can be seen from the air.
There are times when extreme accuracy is required, and there are times when you do not require such expensive accuracy. Our approach is to provide images to those who need to see what is on the ground but are not necessarily making high-accuracy maps from the images. By using sophisticated equipment in the aircraft, we are able to acquire data that allow us to locate and orthorectify the images without the use of ground control points (though if the ground control points already exist, we can also make use of them). The advantage to using this technique is that we are able to offer GIS-ready images that exceed National Mapping Accuracy Standards at the 1:24000 scale.
It is very difficult to put a real number on the accuracy of an image as there are so many variables that affect the final accuracy. Certainly the accuracy of the GPS, IMU, camera, and lens play an important role in the final accuracy. The accuracy of the DEM is the most significant contributor to errors. The DEM defines the ground elevation that is used for orthorectification. If the DEM is in error, it will directly impact both the absolute and relative horizontal accuracy. We use 10-m DEMs and have had good results. Even so, the accuracy of the DEM is our limiting factor. The rest of the system has shown error values that are lower than that injected by the DEMs.
In general, we are finding our images to have an absolute horizontal accuracy of better than 4 m when using 10-m DEMs. We are continuing to optimize the system to improve this number, but we feel that this is accurate enough for many applications.
Realize that the National Mapping Accuracy Standard at the 1:24000 scale for absolute horizontal accuracy is 40 feet, more than 3 times what we are seeing in our processed images. But these are all just numbers. You must determine how accurate you need your images. We can also do better than the 4-m value with ground control points if required. The software we use for orthorectification (Leica Geosystems, ERDAS Orthobase) is capable of processing ground control points to provide very accurate orthorectification.
You must also decide which kind of horizontal accuracy you need. Absolute horizontal accuracy determines how close each pixel is to the actual position on the earth that that pixel represents. It also determines how closely an image will align with other images and with other GIS data.
Relative horizontal accuracy determines how accurate each pixel is relative to the other pixels in the image. The relationship of these pixels determines how accurate a measurement in the same image will be.
Absolute accuracy is not that important if you are just making measurements within a single image. You may be measuring the area of a subject but are not particularly concerned with exactly where the subject is. For example, you may want to know how large a mine has become but have other information about the location of the mine.
It is important that you keep your accuracy requirements realistic, as this has a considerable influence on cost.
Image file format
By GIS-ready, we mean that an image is completely prepared for GIS use. Typically, the image we supply uses a GeoTIFF file format along with a world file that includes all of the data that the GIS needs to locate the image.
Regardless of the GIS you use, the GeoTIFF file format is the most universal as most GIS can read them. The GIS will then convert the image into an internal format and usually pyramid the image to speed the display. However the GIS processes the data, the GeoTIFF format is a lossless format that will not degrade the image. Other file formats, such as JPEG, will compress the image into a smaller file but will result in a lower-quality image. We can provide other image formats if required, but highly recommend GeoTIFF as the most universal and highest quality format for initial image delivery.
Imaging details
Each image we deliver includes the orthorectified image, the world file, the RAW image, and a metadata file that describes the details about the image. The RAW images are like film negatives, not very useful by themselves, but they are the originals and contain all of the image data. The RAW images have a further advantage in that they are 1/3 the size of the orthorectified images. The information within the metadata file includes all of the details for converting the RAW image into an orthorectified image. The smaller RAW images and a copy of the metadata files are stored on a separate CD or DVD from the delivered images as backup.
The images as well as the metadata file are delivered on CD, DVD, or via a portable external hard disk.
The list below describes the information contained in the standard metadata file. Additional data can also be provided as needed. The data in the metadata file allows you to see exactly how the final images were created.
Metadata file
Camera details
Lens details
Camera settings
Camera calibration details
Photoshop processing settings
Gyro coefficients
GPS master station used
Projection details
Datum details
The image acquired is determined by the altitude of the aircraft above the ground, the camera, and the lens used. We use the Canon 1Ds professional digital SLR with a 28- or 50-mm lens. It uses an 11-Mpixel CMOS sensor that is 4064 pixels wide by 2704 pixels high. The lens is decided before the flight so the only option during the flight is the altitude of the aircraft.
As we typically are acquiring images for multiple customers, the altitude above the ground will often be different on the same flight. Our flight-planning software and flight software take this into consideration.
The following table provides information about the images we acquire at different resolutions.
Cm/Pixel |
Horizontal Ground Cover (m) |
Vertical Ground Cover (m) |
Ground Cover (sq km) |
Ground Cover (acres) |
Ground Cover (sq mi) |
20 |
813 |
541 |
0.44 |
109 |
0.17 |
30 |
1219 |
811 |
0.99 |
244 |
0.38 |
40 |
1626 |
1082 |
1.76 |
434 |
0.68 |
50 |
2032 |
1352 |
2.75 |
679 |
1.06 |
60 |
2438 |
1622 |
3.96 |
978 |
1.53 |
70 |
2845 |
1893 |
5.38 |
1331 |
2.08 |
80 |
3251 |
2163 |
7.03 |
1738 |
2.72 |
Customer-supplied information
Before any images can be acquired, certain data is required. We need to know where the site is and what we are taking an image of. We also need to know when you need the images acquired as some images have a window of opportunity for reasons such as the status of leaves on trees. Other details including the map projection and datum will also be required. Over the years, we have found that it is most efficient for all parties if the customer provides us the data we need in an electronic format. This not only speeds the process of entering the data into our image-acquisition database, it also minimizes the opportunity for error (a simple mistake can place a site out in the Pacific Ocean).
The first time you create the information necessary for acquiring images you may find it somewhat confusing. But after the process is worked out (often via email), supplying the data for subsequent images will be much easier.
We are constantly working with our customers to ease the process of specifying image data. Our goal is to make it as easy for you as possible. Typically, a customer will determine which sites are required for the year and make a single submission. Then, we supply the images as we are able. As the images arrive, the customer is instantly able to begin working with them.
The point is, we are constantly working on ways to smooth the process for you to use images. As new projects arrive, we often find it necessary to rewrite some of our software to more easily accommodate the new requirements. We use as much automation as possible to easily accept your data and then supply the images in a way that seamlessly integrates into your workflow.
In most cases, this information already exists within your GIS or other form of database. When this information is not available, we can often help you create it. Realize that to provide these images at the lowest possible price, we must rely upon your data being accurate and requiring a minimum of our time to input the data to our system. Typically, our customers are using some sort of GIS database that allows them to just output this data, usually in a text file. A simple text file is the most universal file format and that is what we typically use. Other formats are also possible, but by staying with the format that is most universal, we can automate the process and keep the price for images down.
The following shows a typical list of the data that must be supplied to us along with some optional data that can ease the acquisition process.
Customer Data
Individual Site ID
Single coordinate near the center of the site OR shape file of the site
Date range when images can be acquired
Elevation near the center of the site (optional)
Type of site
DEM(s) covering the site
Approximate size of the site (optional)
Map projection
Horizontal datum (usually NAD83)
Photo filename convention (usually SiteID_Date)
Minimum solar angle (usually > 45 degrees)
A word of caution regarding map projections and datums. The projection of a 3-dimensional world onto a 2-dimensional surface such as a map is a complex process. Also, the datum used for determining a location on the earth's surface is not always the same. Before we can provide GIS-ready images, we must know what projection and datum you need. It is critical that all of the data used with a GIS use the same projection and datum.
For example, the State of Oregon is using a single projection for the entire state that is often called Oregon Lambert and the 1983 North American Datum. Cities and towns within Oregon may or may not use this projection. The City of Corvallis uses State Plane, Oregon North Zone, and the 1927 North American Datum. There is no real correct choice. You just have to make sure that you are using data with the same projection and datum. If not, the data will have to be reprojected.
Once we receive the data, we will do a quick check that the locations of the sites make sense and then add them to our image-acquisition database. Once the images are in our image-acquisition database, they will be acquired as soon as possible and then delivered after the post-processing is completed.