My last column ("Digital camera technology," May 2006) focused on the impressive capabilities and explosive growth of today's high-end digital mapping cameras. Without question, digital sensors have and are changing the way imagery is acquired for mapping projects around the world.

These devices are also opening new markets, resulting in new and innovative uses of mapping in nontraditional applications, and providing increased capabilities to the end-users of mapping information. As mapping professionals, our knowledge and understanding of the capabilities of these systems is growing exponentially, both in terms of the information found within the imagery and the accuracy that can be expected from map products generated from the digital images.

In my last column I discussed the quality and amount of information that can be gained from today's digital sensors that capture 12 bits of image data. I also discussed the importance of the simultaneous capture of image data in the infrared region for some of today's new mapping applications. Let's now take a look at a few changes brought on by first-generation digital imagery and how these new sensors are being used around the country. We'll also take a close look at a revealing test of the accuracy that can be expected from this new technology.

Digital Imagery Applications

The U.S. Department of Agriculture (USDA) Farm Service Agency (FSA) is a national leader in generating precise image maps of the country. Digital orthophotos for its National Agriculture Imagery Program (NAIP) are generated each year from photography captured during the agriculture growing seasons in the continental United States. Geoff Gabbott, contracting officer for the USDA FSA in its Aerial Photography Field Office in Salt Lake City is one of the program's leaders. He has seen tremendous growth in the use of digital sensors over the last few years. "Film cameras accounted for 90 percent of the imagery acquisition in 2003 as we began the NAIP program. Digital sensors have grown steadily in acquisition for NAIP since 2003. This year, film cameras maintain a slight majority with 51 percent of the imagery acquisition. We expect digital sensors to overtake film in next year's program."

Three different digital sensors will be used for the 2006 NAIP imagery acquisition: the Intergraph Z/I Imaging (Huntsville, Ala.) Digital Mapping Camera (DMC), the Vexcel (Microsoft, Boulder, Colo.) UltraCam D framing camera and the Leica Geosystems (Atlanta, Ga.) ADS40 pushbroom sensor.

Gabbott commented on new applications for digital sensors in the near future: "USDA promotes and advocates partnering with other federal and state agencies. The NAIP imagery is widely used, but many potential partners have needs that are slightly different than what can be served from the 1-meter natural color map product produced under NAIP. Digital sensors have the ability to meet these needs. For example, color infrared imagery is very important to some agencies. The ability to simultaneously collect this infrared information and provide it in a four-band deliverable [along with the red, green and blue found in natural color] has tremendous value in certain applications."

If you're not familiar with the NAIP program, you should take the time to learn more about it. The imagery may have valuable applications in your business practice. For more information regarding the program, go to USDA's website

The St. Louis District of the Army Corps of Engineers is also a national leader in mapping science. This agency provides an important mapping component to the Federal Emergency and Management Agency (FEMA) as well as many other federal, state and local governmental agencies, especially in times of natural disasters like last year's hurricanes. The Corps has chosen to use digital imagery from Leica's ADS40 digital sensor during these times because of the ability to provide 24-hour turnaround of image mapping. Dennis Morgan, chief of the district's Geospatial Engineering Branch emphasized the importance of the quick turnaround in times of a national disaster: "The rapid-response turnaround is critically important to those in the field assessing the damage after hurricanes. Film is no longer a viable option for these projects due to the one to two days that are added to the product cycle from the film processing and image scanning-steps that are not required with first-generation digital imagery."

Morgan sees many other opportunities for digital image acquisition with the district's clients. "The ability for rapid turnaround and the improved image resolution found in today's digital sensors are impressive. But the added ability to capture infrared imagery [in addition to panchromatic or true color data] simultaneously, and the ability to acquire quality imagery at lesser sun angles can be equally important for the right projects."

Today's digital sensors, like the Vexcel UltraCam D, are allowing mappers to immediately produce maps after landing, and providing them with quality imagery as well as the ability to map heavy terrain with great accuracy over their film counterparts.

Accuracy Analysis

We are relatively early on in the technology curve with high-end digital sensors. Because of this, there is limited historical information regarding the accuracy of planimetric and topographic features captured from first-generation digital sensors. This is significantly different from film cameras where the accuracy of map features has been well-documented over the last several decades.

The Pennsylvania Department of Transportation (PennDOT) was very interested in taking advantage of the capabilities of the Intergraph Z/I Imaging DMC framing camera for its transportation mapping. But it wanted to develop a clear understanding of what accuracies could be expected from this platform before moving forward. Michael Loose, photogrammetry manager for PennDOT, explains: "We had no doubt the DMC would be as accurate as film. But we wanted to see if we could fly higher and maintain the accuracy that we require in our mapping. If so, it would help reduce the control and mapping costs for our projects."

We at Photo Science had performed several less-formal accuracy analyses of the DMC, but were interested in working with PennDOT in a formal test of the DMC in large-scale mapping. The tests were designed to include a comparison of accuracies gained from traditional film photography captured at two different elevations. The analysis began with flights over State Route 924 near Frackville, Pa., in April 2005 with the DMC at an altitude of 2,000 ft above ground, and the first of two film flights at an altitude of 1,200 ft above ground. (Altitudes of 1,200 to 1,500 ft are fairly common in transportation mapping.) The second film flight was conducted in November 2005 at an altitude of 2,000 ft above ground. All film photography was captured with a Leica RC30 precision photogrammetric camera.

The flight parameters were slightly different for each of the three flights due to the varying altitudes and the smaller footprint of the DMC when compared to film at the same altitude. The table on page 33 illustrates the information for each of the flights.

In mapping, accuracy is highly correlated to flying height. Lower flying heights produce higher accuracies when compared to flights at higher altitudes. Therefore, data from the two flight altitudes were carefully evaluated in the tests along with the comparisons among platforms. To ensure a statistically valid test, all three sets of imagery were oriented for map compilation using the same ground control points.

Forty-five checkpoints were surveyed for the purpose of testing the three sets of vector mapping produced from the three independent imagery sources. Differential leveling was used to determine the elevation of all checkpoints to gain a high elevation accuracy. Fast Static GPS surveying methods were used to determine the horizontal position of 20 of these points. The checkpoints were distributed over a diverse area.

The accuracy of the DMC was compared to similar mapping produced from traditional film. And the results were revealing. The table above lists the statistical results of this evaluation and compares the results against the requirements of the ASPRS Class 1 accuracy standards (in terms of the root mean square error, or RMSE) for mapping compiled at a map scale of 1''= 50' with a 1-foot contour interval.

The first thing that should be clearly evident is that the mapping products from all three flights generally meet the ASPRS requirements (the RMSE for the easting for the higher film flight is slightly higher than allowed). But the real significance here is that the DMC significantly outperformed film in terms of the horizontal accuracy and was equal to film in the vertical accuracy.

PennDOT's Loose was pleased with the results of this test. "Everyone [who] worked on this project on our end was very impressed with the quality of the imagery. Edge detection was much better than any film projects we have worked on. Also, the ability to see in shadows (due to the added information found in the 12-bit imagery) is a significant advantage to us, especially given the rugged terrain in our state and the shadows common in this terrain. We plan to move to a DMC platform as soon as possible."

Digital's Many Advantages

The new digital sensors offered today are impressive in traditional mapping applications, and are opening doors for new and innovative ways to use mapping information. They provide mapping professionals the ability to begin map production immediately after landing the aircraft in time-critical applications. They provide an image quality never before seen in the mapping environment and allow quality image acquisition in less than ideal environmental conditions. And they are proving to be even more accurate than their film counterparts. The future is indeed bright for this new technology.

To read Mark's previous articles, click to and search the Point of Beginning database for "Mark Meade" or "From the Ground Up."