Introduction
It is inevitable that everyone has something to hide. This includes even the most honest of people. Whether it is that they cheated on their diet, to the candy bar that they stole when they were six, people have secrets that they either need or want to keep. The rationale for keeping secrets isn’t always so clear or even logical. Some of these secrets are exposed by taking the higher ground. By taking the higher ground and using some form of aerial imagery, it is possible to expose some of these secrets, whether dirty or innocent. In the case of the moonshine distiller or the marijuana grower, the reason to hide this secret is easy; getting caught is bad and cuts into their profit margin. In the case of a state or government, the reasons may not always be so clear. Generally the reasons usually have to do with national security, and the general welfare of the people in the care of the state. However, the successful grower or distiller will find or create methods to conceal their labors, through methods of camouflage, misdirection and other creatively available methods. Governments, in an attempt to conceal sensitive information, will also do the same. It is herein that the question lies: what am really I looking at?
“Ground Truthing”
The easiest way to confirm an aerial photo is to physically visit the location, if at all possible. This method of verification is loosely termed ground truthing. Ground truthing carries a couple of definitions. However the basic idea of ground truthing is to verify an image is accurate in its depiction of the terrain and to verify that no anomalies have occurred. In an extreme sense, ground truthing can also be used by strategic planners to ensure that a military asset is what it is meant to be. An easy but good example of this occurred during World War II and the ensuing years after the war. In order to trick the enemy, the allies would place dummy tanks complete with dummy tank tracks to fool the photo interpreter (see fig. 1).
Fig. 1
The tools often used for this type of campaign generally relied upon deception, misdirection and camouflage. In a non-military environment, a good reason to ground truth an image is that an image may contain some unknown variable. However, without visiting the site, the image interpreter may be unable to accurately discern certain features such as a house that may be hidden within a grove of trees. In ground truthing, there also may be errors in the imagery. Some of these errors can be classified as errors of omission or commission. An error of omission may occur when an object is not classified, even though there is a classification for that object. For instance, though a manhole may be visible on an aerial, the interpreter of the aerial image may fail to classify it as a manhole. An error of commission may occur when an object becomes classified as another object that it is really not. An example of this would be an asphalt road labeled as a train track. (Wikipedia; Ground Truth). It is important to ground truth because remote imagery has many limitations.
Limitations in Collection and Production
In aerial mapping and remote sensing, there are many limitations which can affect the final outcome of the product. Some of these limitations are directly related to the sensor, the airborne platform, the environment, the interpreter/user of the information, and multiple other factors such as cost. In order to mitigate these effects, there are a variety of solutions which may be employed.
The biggest cause for error is human. Being human is in itself a limitation that enables us to do or not do certain tasks well. It is for this reason that one of the limitations if imagery is human error. In classifying objects in an image, the image is only as good as the interpreter. As discussed previously, this is one of many reasons why we ground truth. Although we are human and this is a limitation in itself, we are also limited by how well we understand the technology, our ability to exploit the technology, and the ability of the technology to adequately perform the required task.
One of the most obvious limitations in acquiring imagery is the weather. Weather factors such as cloud cover, fog or snow may make it difficult to obtain quality images for both satellites and aerial platforms. Additionally, weather factors such as wind may make it difficult for an aircraft to fly within tolerance of the planned flight path required to obtain quality data. Even though weather has its limiting factors, there are ways to overcome this. Sensors attached to a space based platform are rarely affected by weather. Radar sensors such as Synthetic Aperture Radar (SAR) and side looking airborne radar (SLAR) have the ability to penetrate clouds. In general, space based platforms are able to outperform the aerial platform as the space platform is not weather dependant.
Trees and other ground features may also create limitations in obtaining quality imagery data. Trees and shadows from trees and other objects may make it difficult to see objects on the ground. As discussed previously, radar sensors have the ability to penetrate clouds. Unfortunately, radar sensors consistently fail to penetrate through dense objects such as forests. This makes it difficult to acquire good data about ground topography and ground features from radar in forested areas. The best way to verify objects on the ground is through ground truthing and supplementary ground survey.
When acquiring aerial imagery, the cost and time of travel must also be taken into account. It is possible for some companies to affix multiple sensors at one time. As technology becomes better, some sensors have incorporated multiple modes of acquisition. An example of this would be an operator such as Keystone aerial surveys out of North Philadelphia (KPNE) who use the UltracamX. “The UltraCamX is a multispectral imaging sensor capable of capturing panchromatic, red, green, blue, and near infrared data simultaneously” by incorporating multiple sensors, the operator has the ability to better serve the client through faster turnaround times.
Another consideration is the need to set up control points on the ground for aerial surveying. These control points on the ground provide a way to correlate the images collected with the ground grid coordinates and elevations. Setting up control points is also a form of ground truthing. Setting up ground control points is a project cost and requires additional time and coordination. It requires a survey crew to go out and set up panels, and survey the panels location. A process known as analytical triangulation is used to correlate the ground control points with the aerial data. In the article titled “From the Ground Up: Direct Georeferencing in Aerial Photography” by Mark E. Meade, the author states that one way to overcome the need to set up control points is to use Airborne Global Positioning Systems (ABGPS). By using ABGPS, it is possible to know the aircrafts position relative to the earth at all times. With the advent of Airborne Global Positioning Systems (ABGPS), the number of control points required for a project is greatly reduced, thus saving time and money.
There are different advantages and disadvantages to the use of satellites versus an aerial platform. Satellites are generally more able to acquire large areas with relative ease, whereas an airplane may require several flight lines and multiple control points in which to gather the information. Aerial platform radar systems tend to be a good method to map an area quickly and efficiently with a relatively lower cost, because the flight line tends to be wider and would require fewer flight lines flown In this same way that radars are used on aircraft, this is a sensor that also translates to a space-borne platform.
A satellite can often produce the same image that an aerial mission would acquire, but would be at a greater cost. For small projects, satellite may not be cost competitive with the aerial platform. For large projects, the time saving benefits of using satellite might outweigh the additional cost.
Let’s just propose an example. In this example I need to do a noise abatement study because commercial traffic at Prescott Love Field (KPRC) in the last two years has increased by 15%. The surrounding neighbors keep calling the City of Prescott airport manager and complaining about how noisy traffic at the airport has become and how they can no longer sleep at night. It is bad business to drive this traffic away, since this is the very traffic that is bringing money in for the city. In an effort to solve this problem, the city manager along with various other authorities would have to come up with a way to move the traffic away from the housing development. This would keep the airport manager employed and the neighbors happy. In order to effectuate this plan however, an aerial image of some sort must be acquired in order to show any new development that may have taken place. The City must minimize the cost required to perform this study.
Aerial imagery of this nature is readily available through various vendors. One of the more prominent online vendors, mapmart of mapmart.com prices imagery with a two foot resolution for satellite at $822. This same area can be acquired for $567 if you desire imagery from an airborne flight. Although this is a savings of $255 (even for large companies that is a decent savings), the imagery is slightly outdated (See Fig. 4 below).
Fig. 4
It would likely cost even more money to have aerial photographs of this area taken specifically for this project. The problem is that the City must balance the cost of the data they need against the quality of the data. If the Airport Manager believes that growth to the area surrounding the airport has been minimal over the few years, the older data may be acceptable. If the Airport Manager is unsure of how much growth has occurred recently, or if recent growth has been substantial, the manager may choose to have aerial imagery collected specifically for this project, despite the increased cost.
Conclusion
Though there are limitations to all forms of imagery data collection, satellites generally seem to have fewer limitations. There are still applications in which data from an aerial platform may be preferable to data from a satellite. Because both forms of imagery are imperfect, the only way to know for sure what is being seen is to ground truth the data. Through real time verification, we can ascertain the properties of hidden marijuana plants or mislabeled tanks with the all inclusive fake tank tracks.
For a complete works cited please leave a comment.