Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Photogrammetnc Measurement and Mon~tonng of Histonc and Prehistonc Structures THOMAS R. LYONS and NAMES I. EBERT Photogrammetry is measurement using photographs or other images of re- motely sensed electromagnetic data. It provides~architects, archeologists, and others concerned with prehistoric and historic cultural resources with cost- effective means of recording and monitoring structures, monuments, and other cultural sites. This presentation sets forth the basic principles of photograrn- metry, concentrating on the use of vertical aerial photographs and terrestrial recording of structures with a ground-based camera. Controlled stereo pho- tographs can be used as the basis of a number of photogrammetric products, including photogrammetric maps, orthophotos, and combinations thereof. Re- petitive aerial and terrestrial photographs of a site, and products derived from them, can be compared to allow the monitoring and documentation of inev- itable change in cultural resources. Methods by which this is accomplished are illustrated by reference to a number of ongoing monitoring experiments. Considerations dictating frequency of monitoring and precision of measure- ment are discussed, as is the multiuse nature of photogrammetric products. A wide-spectrum approach to monitoring cultural resources that integrates the use of past data sources such as historic photographs and maps, present- day photogrammetric products and data, and possible future techniques such as microwave scanning and holography, is necessary to ensure the conservation and integrity of our historic cultural resources. -Thomas R. Lyons is Chief, Remote Sens~ng Division, Southwest Cultural Resources Center, National Park Service, Albuquerque, New Mexico; Jazzes I. Ebert is Archeol- ogist, Southwest Cultural Resources Center. 242
Photogrammetric Measurement and Monitoring of Structures 243 This paper is concerned with the methods of aerial and terrestrial photogrammetry and their application to the measurement and mon- itoring of historic and prehistoric architectural structures. The purpose of these procedures for documenting the condition of building fabric is to provide a basis for quantitative measures of alteration or deteri- oration over time and thus to provide data for informed decisions on maintenance, preservation, or restoration. The photographic products of calibrated aerial and terrestrial cam- eras are historic documents of value to historical architects and/or archeologists. Under certain circumstances of acquisition they provide a data base that can be used for comparative purposes, for faithful restoration or reproductions.) - Repetitive controlled imagery has the added function of establishing a measure of the rate of change in building fabric and consequently a foundation for predicting future deterioration. It therefore allows the foreknowledge necessary to prevent or mitigate harmful effects. INTEGRITY OP HISTORIC STRUCTURES The value as heritage and the scientific significance of historic and prehistoric architectural structures depend to a great extent on the integrity of the fabric of which they are composed. The term "integrity" as used here does not imply the absence of change. Change in materials over time is inevitable, and humanly induced changes in structures occur regularly and can add interest and value to historic properties. A historic structure has integrity when the process of change through time can be documented and when the nature of the changes and their causes have been determined. Photogrammetry, a remote sensing tech- nique involving accurate measurement from photographs or other im- ages of electromagnetic data, is a powerful too! for the historian, his- torical architect, or archeologist wishing to document the history and causes of change in architectural resources. BASIC PRINCIPLES OF PHOTOGRAMMETRY Photogrammetry Photogrammetry, in its strictest sense, is measurement using photo- graphs. A single aerial photograph for instance, is a representation of a three-dimensional scene reduced to a two-dimensional format. While it fairly presents scaled measurements for each plane in the scene parallel to the film plane, it also contains a number of distortions. One
244 CONSERVATION OF HISTORIC STONE BUILDINGS of these is occasioned by the fact that the relationship between the actual size of an object photographed and its size on the film depends on its distance from the camera (focal length of the lens being held equal). Even in a vertical aerial photograph, this causes differences in scale between an object on the top of a mountain and an object at the bottom of a valley since they are different distances from the camera.2 Radial or relief displacement is another effect of forcing a three-di- mensional scene onto a planar film. For example, if a tall object such as a masonry obelisk appears at the exact center of a vertical aerial photograph, only its top con be seen; but if it appears anywhere else in the image, not only its top but also one of its sides may be seen (Figure 11. A single, monoscopic photograph containing-such distor- tions obviously is not an accurate representation of reality. Cameras used in photograrnmetry are classified by the platform on which they are placed. Space or airborne cam eras are aerial cameras, while those resting on the earth and pointed horizontally rather than vertically are terrestrial cameras. The difference in the orientation of these two types of cameras allows the architect and archeologist to derive accurate representations of structural elevations and plans from both a vertical and horizontal perspective. Stereophotography The use of stereophotography turns the distortions inherent in mon- oscopic photos into advantageous information.2 Two sorts of mea- surements that can be made from a stereo pair of aerial photographs are parallax measurement and radial line plotting. Parallax measurement makes use of the radial displacement inherent in each of the photographs in a stereo pair. The difference in distance between the base of an object and the top of the same object in each pair is measured; if the scale of the photographs is known, the height of the object measured can be determined. The principle allows the measurement of the relative heights (z coordinates) of each point in a photograph and is the basis of topographic mapping. The accuracy of the x, y, and z dimensions derived through parallax measurement and radial line plotting from a stereo pair depends on the quality of the photographs themselves. The film plane must be extremely flat to minimize uncontrolled scale variations within the image, and the lens axis must be accurately perpendicular to and cen- tered over the film plane. As a consequence, photographs taken with standard cameras are not generally suitable for mapping, and metric cameras are used instead. Unlike "snapshot" or even higher-priced StR
Photogrammetr~c Measurement and Mon~tonng of Structures I..: ~ - r..~ . ~ :: ~~:::~ FIGURE 1 Relief displacement illustrated by an aerial photograph showing an obelisk at Chalmette National Historical Park in Louisiana (upper left coiners. If this obelisk had appeared at the exact center of the photograph, only its top would have been visible. Since it was at the edge of the photograph instead, the sides are clearly visible. cameras, metric cameras are designed to take extremely accurate pho- tographs. They are collimated to insure that the axis of their lens is perfectly perpendicular to the film plane, their platens bear fiducial marks so the exact center of the lens axis can be determined on the photograph, and they are precisely calibrated in a laboratory prior to sale. While metric cameras range in price from $10,000 to $60,000 or more, a number of firms maintain and use them regularly and will enter into contracts for terrestrial and aerial photogrammetry. True horizontal mapping of structures or features from vertical aerial photographs is accomplished through radial line plotting. Radial line
246 CONSERVATION OF HISTORIC STONE BUILDINGS plotting makes use of the geometric fact that any point in space can be defined by sighting it from two known reference points or stations and recording the angles to the unknown point. In mapping, this pro- cecture is referred to as triangulation. Radial line plotting using a stereo pair of photographs is usually done with a radial plotter or photogram- metric plotter. The instrument uses the center of each vertical photo as a known station and allows the interpreter to position a line from this point through each point of interest on the photograph. Such points might trace the outline of a structure or site feature. A mechanical linkage provides a means of moving a tracing pencil, which creates, in effect, a matrix of points of intersection that planimetrically rep- resents the location of features on the ground.3 Although such data can be obtained through a field survey using a transit or alidade, radial line plotting from aerial photos is many times faster and more eco- nomical. Accuracy, Precision, and Scale Accuracy is the closeness of a measurement to the real world; it de- pends on the scale at which photographs are taken and interpreted and the scale at which these measurements are plotted. In general, the larger the scale of a photo negative, the more accurate a final photo- grammetric product will be. Precision refers to the distribution of the values of a number of samples of a single parameter about the mean value in other words, the replicability of a measurement. Precision in photogrammetry depends primarily on distortions or errors asso- ciated with the photography, the plotting device used, and the plotter operator. These errors in turn depend to a certain extent on the scale of imagery end map plotting, especially the variation from one inter- pretation of a stereo pair to the next by the plotter operator. Accuracy is important in photogrammetric mapping; accuracy and precision are vitally important in the sorts of comparisons that constitute photo- grammetric monitoring of architectural targets. The scale at which controlled photographs of a site or structure are taken must be decided on the basis of a balance between the constraints imposed by reality t economies and the capabilities of equipment) and the needs of the researcher (problem orientation). It is not practical, nor is it very definitive, to ask for "as large a scale as possible," and the scientist or manager who does so probably has not thought seri- ously about his need. If photogrammetric products and data are to be used for accurate reconstruction of a structure, the accuracy to which building materials can be fabricated will be a criterion for detemi~ning scale. When a
Photogrammetr~c Measurement and Monitoring of Structures 247 structure is being recorded for monitoring purposes, questions of how much its fabric changes cyclically {diurnally, for instanced and what sort of change would signal impending collapse or irreparable damage must be asked.- Answering these questions may require relatively high accuracy. Theoretical or scientific questions, on the other hand, may not be so demanding. If one were to survey the ethnoarcheological literature for predictive formulae for determining the population of structures, for instance, one would find that the variance in estimates and observations is high enough to make the use of any photogram- metry at all questionable. Photogrammetry need be no more accurate than the problems to which it is to be applied warrant. In applying this rule to the deter- mination of scales, it should be remembered that to double the ac- curacy at which a map can be plotted, one must double the photo- graphic scale and therefore quadruple the area that must be photographed and interpreted. Necessarily, there is a corresponding increase in cost. Precision- or replicability of photogrammetric results, which is nec- essary for comparing periodic data from a specific site or structure, is a problem. By permanently- marking and reusing control points and employing the same metric cameras, camera positions (difficult from the air), and plotting equipment, a certain amount of comparability between maps can be ensured. Another problematic element is inconsistency on the part of the plotter operator. Photogrammetric maps are interpretations of photo- graphs, accurate ones to be sure, but including different amounts of detail depending on the inclination of the human interpreter. This is especially apparent with contour lines drawn on two maps of the same target. Topographic contour lines are very difficult for the same op- erator, using the same imagery, to duplicate exactly. Point locations, on the other hand, can be duplicated by different interpreters with surprising accuracy. For this reason, it is advantageous to locate specific points to be used as the basis of monitoring a structure rather than attempt to use a contour map of, say, the face of a wall as monitoring data. This requires the selection of conspicuous, relocatable points (along the top of a wall, at the corners of windows, etc.) and may require that permanent markers be placed at these points for some structures. Orthophotography An orthophoto is essentially a photographic planimetric map. It is produced by a machine that not only does radial line plotting, but also uses parallax measurement as the basis of changing the scale of very
248 CONSERVATION OF HISTORIC STONE BUILDINGS small segments of a monoscopic photograph as these small segments are exposed on a film. This corrects for scale differences throughout the photograph and creates a planimetrically correct print of a scene. Orthophotos are more useful than planimetric maps for many purposes because there has been no selection by an interpreter of what points in the scene are of interest; all points Within the limits of resolution of the photograph) have been shown, and details not available on a map can be distinguished at a later date. A combination of an orthophoto and a photogrammetric line map can be made by first compiling an orthophoto and then using the same controlled stereo pairts) on a photogrammetric plotter to produce a topographic contour map at a desirable scale. The line map is then photographically overlain on the orthophoto (in negative form white map lines against the darker orthophoto, Figure 21. Such a product combines the advantages of both of its parts: Lan~narks and other details can be found and used for measurement on the orthophoto, while quantitative, three-dimensional data are available from the top- ographic map lines. An orthophoto/topographic map combination is especially useful as an aid to plarlIiing survey, excavation, or conser- vation activities in the field. Ground Control To relate the three-dimensional coordinate values of each point pho- togrammetrically measured in a stereo mode! to other measurements from the real world, it is necessary to mark and measure a few control points on the ground or target to be photographed before the photos are taken. Ideally, three or more horizontal and four or more vertical control points are set and marked so that they will be visible in both of the photos forming a stereo pair (Figure 31. The distance between the horizontal control points is measured, and the difference in ele- vation of the vertical control points is established. This allows the operator of the photogrammetric plotter to determine the precise scale of the photos and insert them (printed on glass plates rather than paper, and called diapositives) in the plotting machine. If control is not set prior to taking stereo photographs, certain conspicuous points on the ground can often be "photoidentified" in the image and then located and measured on the ground, serving as control points after the fact (Figure 41.
Photogrammetric Measurement and Monitoring of Structures 249 _ - ~ _ ~ _ ~ at_ ~ _=_ ~ _ me_ - __ ~ __ _~ _ _~ FIGURE 2 An orthophotograph of a Pueblo-m Period masonry site in San Juan County New Mexico, with superimposed topographic contours. The orthophotograph is essen- tially an aerial photo corrected to show a planimetric view of the site while also con- veying details inherent in a photograph. Topographic contours were compiled using a stereo plotter and photographically superimposed over the orthophoto, thus supplying metric information as well. Such renditions are especially useful as base maps for plan- ning and executing fieldwork. SOURCE: Koogle and Pouls Engineering, Inc., Albuquerque, New Mexico.
250 CONSERVATION OF HISTORIC STONE BUILDINGS _ _ 1 1 Area of ~ Stereo Coverage FIGURE 3 Minimal horizontal {triangles) and vertical ;circlesJ control points required to be marked and measured in a stereo model. When flight lines are being used as the basis of mapping, it is sometimes possible to use fewer control points and still maintain photogrammetric control by "bridging.". Control points are located in the field prior to flying and are marked with stakes and plastic panels so that they are visible in the photographs taken later. The distances between horizontal control points, and the dif- ferences in elevation between vertical control points, are then measured. Control points can also be tied to known data points if desired. Aenal (Far-Range, vs. Terrestnal (Close-Range, Photogrammetry Aerial photographs are the staple for far-range photogrammetry. -They are used every day by engineers, planners, geologists, cartographers, and many others for the development of topographic and planimetric data. Aerial photographs are taken in scales ranging from about 1:200,000 (small-scale photography) to 1:500 (large-scale photography). Equipped
Photogrammetr~c Measurement and Monitoring of Structures 251 _. FIGURE 4 An aerial photograph of the Barbourville Mansion in Virginia, a brick struc- ture designed by Thomas Jefferson and destroyed by fire in 1844. The L-shaped white markers in the corners of this scene are ground control panels, set prior to overflight. If these panels had not been set, it would have been necessary to photoidentify points and measure the distances between them later. Examples of points that are well defined and could be used for "subsequent control" are shown at A and B. with the most commonly used aerial cameras, airplanes cannot fly slowly and low enough to take larger images. Using a first-order plotter, a photogrammetric map scale of 1:100 (1 in. = 8.33 it) can be obtained using 1:1,000 scale imagery at a contour interval of as little as 0.1 ft t3 cm). The use of aerial photography for photogramrnetry offers the possibility of mapping very large areas inexpensively but at low res- olution and smaller areas more precisely up to this limit. Because of the design of most aerial plotters, oblique aerial photography is of little utility in photogrammetry.
252 ~ ._ o . CO ~ sip o ~ o ~ ._ ~ C) :> {~, an ~ _ - ~ c ~ .~ ~ _ ._ _ au ·= N ~ O ~ Cal V O . - O . - O ~ _ .0 ~ ·m V ,.~ : C) ~ ~ V ~ - I) ~ V ~ ~ to O ~ _ N V O ~~ ~ O ~ O V en - C) ~ o s=¢ V m- ~ O-S: Lr' ~ ~ v m- O .0 ~ ,~ ~ ~
Photogrammetr~c Measurement and Monitoring of Structures 253 Terrestrial photogrammetry uses photographs taken with a terres- trial photogramrnetric camera or a phototheodolite, the ground-based equivalents of an aerial camera. For terrestrial photographs, the film plane of the camera is arranged to be parallel with the structure being photographed. Control is as important in terrestrial photogrammetry as it is in measurement from stereo vertical photos. Terrestrial pho- togramrnetry results in maps or drawings showing planimetric detail in the x and z (or y and z) planes, with topographic point or contour data in the y or x plane {Figure 51. Most terrestrial photogrammetry in the United States is done with photogrammetric plotters designed primarily for use with vertical aer- ial photos. There is little latitude for image tilt or scale differential with such machines, making camera setup very critical. Some Euro- pean plotters are designed for use with terrestrial stereo photographs, however, and allow the plotting of planimetric detail from oblique stereo photos. These instruments have been coming into use in the United States in recent years. They can transform terrestrial photo- graphs into a "vertical," and since they obtain imagery at much larger scales than possible from an aircraft platform, they can be used to advantage in large-scale drawings of historic and prehistoric architec- tural features. Digitizing Photogrammetr~c Data The products and representations of photogrammetric data discussed above have all been graphic. Another useful way to represent photo- grammetric data is by digitizing three-dimensional point coordinates. Many modern plotting devices incorporate a digitizer that records co- ordinates chosen by the plotter operator on magnetic tape or computer cards for later analysis and manipulation. Points may be chosen at regular intervals along the tops of walls or the outer perimeter of a building {from vertical, etc.), or at intervals on the surface of a wall or structural member (with terrestrial photographs). The operator may also become an interpreter, when properly trained, and identify only those areas of significant change in, say, the elevation of the top of a wall, that should be measured. Figure 6 is an aerial photogrammetric map of a prehistoric pueblo structure at Chaco Canyon National Mon- ument with x's marking some of the points digitized during photo- grammetric plotting. Digital photogrammetric data can be used in a number of ways in structural documentation. The simplest of these is graphic represen- t\ation of a structure as a matrix of points connected by lines. Such a
254 CONSERVATION OF HISTORIC STONE BUILDINGS ~~ Fuji ,' ~~ 606155~ ~~ 51~ ~ ~65 ,~6o60i ~ I__ I'd\ ~ ~ m) ~ 1 t°~° '\\ If 'at ''L~ l ~ x Spot Elevations w/o First Two Digits Contour Interval = I Foot Datum is Mean Sea Level '_~ an' red ~: 042.05 Lit FIGURE 6 Photogrammetric map of Kin Bineola Pueblo, Chaco Canyon National Mon- ument, New Mexico. The elevation figures on room floors and along the irregular tops of walls Numbers not shown) were digitally recorded for computer manipulation. SOURCE: Drafted from photogramrnetric data by Jerry Livingstone, Southwest Cultural Resources Center, National Park Service.
Photogrammetric Measurement and Monitoring of Structures 255 representation can be rotated automatically by a computer to create views from any direction or angle or elevation (Figures 7 and 81. Dig- itized data from a drawing of one map or elevation can be compared with subsequent data as the basis of monitoring a structure. Engineering and Art lust as different scales, and hence accuracy and precision, are acceptable or required for different needs and problems, the form of the final product produced through photogrammetry may vary according to the applications to which it is to be put. One of the most often noticed differences is the level of "artistic refinement" of terrestrial photo- grammetric elevations of structures. The first product to emerge from a photogrammetric plotter is a manuscript map or drawing. It shows all planimetric detail and three-dimensional coordinates or contours chosen for interpretation, but often is drawn by a pencil or single- width plotter pen. The lines on some manuscript maps are thus often ~- FIGURE 7 A perspective line drawing of Kin Beineola Ruin produced by a computer from digitized wall height data. It should be noted that "hidden lines" have been shown in this version, conveying more information than if they had been hidden, but also tending to confuse casual viewers. Computer programs exist by which such hidden lines can be removed.
256 CONSERVATION OF HISTORIC STONE BUILDINGS Of ,~N,,~ rat a: -~m rid FIGURE 8 A computer perspective line drawing of Kin Bineola Ruin from a different angle. When combined with room-floor data that indicate the depth of rubble fill, such perspective information can serve as the basis of computer reconstruction of a structure See page 2691. thin, difficult to reproduce, and far from aesthetically pleasing (Figure 91. Photogrammetric firms usually compile a second product from the manuscript map by drafting or scribing the lines represented on the manuscript map for greater clarity or reproduction. Drafted maps contain the same information as manuscript maps and are far more useful for publication. Interestingly, during some early experiments in mapping and monitoring structures at the Remote Sensing Division of the Southwest Cultural Resources Center it be- came apparent that even drafted maps and drawings were not accept- able, or at least not very exciting, to architects. For their purposes, a third-stage product is necessary, and for want of a better term we shall call it an "artistic rendition." Such a drawing contains not only pho- togrammetric data and cartographic information, but information about how a structure actually looks. Unlike a photogrammetric map over- lain on an orthophoto, such information is often idealized, conven- tionalized, and selectively added. Two elevation drawings of the same structure {Figures 10 and 11) serve to illustrate the differences between such products.
Photogrammetnc Measurement and Monitonng of Structures 257 ,. \\ ~~ A. a a _ ~ b ~ ~ 4' ii Am- ~ ~ ~ 5~\~\\~ I~ 1 at? but it -- 1 . 1.~ FIGURE 9 Manuscript map of a portion of a histonc, abandoned Hidatsa Village at Knife River Indian Villages National Histonc Site in North Dakota. Ike round, depressed areas are the outlines of excavations over which earth lodges were built by the Indians. Note the unsharp and difficult-to-reproduce lines, typical of the first stage of photo- gram~netric plotting done with a pencil or similar instrument fitted into a servo-mech- anism connected with a photogrammetric plotter. SOURCE: Koogle and Pouls Engineering, Inc., Albuquerque, New Mexico. THE ECONOMICS OF ARCHITECTURAL PHOTOGRAMMETRY Cost Effectiveness The size and detail of the target of aerial photogrammetnc mapping are the prime factors in determining cost effectiveness. Aircraft setup time, ferry time, time over target, setting of ground control, film and film processing, and laboratory and plotter time are all cost elements
258 ~ 2 - ' t~ 3= of, U 11-1~' 1 ,, to 41;~= '5~ oil | ~ ~ ~ o ~ ~ 3 o; ~ ~ ° ~ CL) o ~ ·4., C) 'A [Z~] _ C,0 ~ ° o
259 ~ ~ cn ~ Cot ~ o A) V: ~ o 3 ~ Ad ·- o ~ ~ o be _ .= ~ ~ ~ ~ .cn ~ _ O ~ ° C'¢ ~ · _ ~ ~ ° ~ MIX ° ~ 'n _ o 3 ~ . - 1 ^. 1 I, 5 1 ~ l l . 1 l j a,. ; 1 ,. i . · V 1 '\ ~ ~ ~? ~ '; , ~ 1 . . . ,. , ~ ,. ,. . . 2 I i ", . t, ._ ~ _~ F' Em ~ I '1 '1 ~ is 1 I! 1 ~ al t ~ ! ~ . ~ r _
260 s In
261 .s 4 - o - C~ .° ~ o me, Q ~ 4~ em o . ~ _. - o o - ~4 o em .C)] lo, C) . ~ Ct o ~ be o ~ · _4 ·_ 3- ~ o o Cal .o ~ 3 ~ 5 or .s 5 . . o . o ._ Cal x o o U' em CO 4~ CC ~4 4= In
262 ~\r) 1 - ~: 5 =~,1~1~1~1~\\\\ \ :~ it) ) -~ Wa%,~N :- ~F ~ Q) _ · ~ UJ 6 C'
263 ~ 3 ~ ID C) al Z V ·V ~ ~ V au it.= o ~ lo. - ~ .m ~ _ ~ Cq (L) ~ en ~ ~ ~ TV ~ v ~ X 3 V .= - A: .= ~ ~ o o ~ ~ c., V o ~ ~ ~ ·_ ·_ Cal o ~ ~ ~ ~ ~ D .~= O O at, O V X As, O ~ ·_ Cal CC O O ·= o 3 ~ Z <I) it, ° <`s o- ~ ~ - - O ~ ~ ~ D Co ~ In P4 ~ V - · t1S ", , i ~0 ~ e~ ~ _ a' .~-asc~ ;> C.) o (O ,= ·s ~ =m $ O CO 4~ 0 U. V cd ~ ~ ~ ~ 8 ~ ~ . . ~: o V) o ._
264 CONSERVATION OF HISTORIC STONE BUILDINGS to be considered. In general the larger the target or the number of targets covered during a single mission, the more cost effective the mapping method. It is much more difficult to determine the cost effectiveness of close- range photogrammetry. Again, however, the more the required detail, the more apt the method is to provide savings over manual or standard surveying and recording techniques. APPLICATIONS RESEARCH Monitoring Photogrammetry, as illustrated in this paper, has proved an efficient means of accurately recording architectural data and is being used experimentally for monitoring historic and prehistoric sites and struc- tures. We would like to suggest, however, that photogrammetry in its strictest sensethe use of controlled photography taken with metric cameras and converted maps and elevation drawings or other photo- grammetric productswill be supplemented in the future by a number of other remote-sensing measurement methods. Because structural monitoring is a historic process, entailing the use of previously col- lected baseline data and comparison with subsequent data sets, it is vitally important that future means of recording and monitoring be compatible with past photogrammetric data bases. The first step in this process was initiated at Pueblo Alto in Chaco Canyon National Monument, a ruin of the Pueblo-III period occupied during the eleventh and twelfth centuries A.D. After the masonry walls in Pueblo Alto were stripped of their overburden of sand and building debris, controlled aerial photographs were taken and a detailed map of the exposed structure was plotted Figure 121. Following excavation and the exposure of additional ar- chitectural features and after reidentifying and panelling of the original ground control, the site was again overflown. This imagery was placed into the plotter and projected onto the original map. It was not nec- essary to remap totally the target but simply to add new details or modify old ones (Figure 131. This additive process suggested the feasibility of projecting past base- line data onto the original photogrammetric map or drawing and check- ing for any changes or modifications to the original draft. Experiments have begin to determine the type, degree, and frequency of change due to natural or human impact on fragile masonry structures. Plans and elevations of historical Barbours Mansion, designed by Thomas Jefferson and burned in the mid-1800s, and of the prehistoric Tower
Photogrammetric Measurement and Monitoring of Structures 265 : ~ . _ _ ~ x`, ~.r 'go L go, 1 ~ ~ x., .. . . ~ rip ~ X ~ ~ \ ~ ~ S r it_ ~ ~ Fir- ° ~ C,= As, ,~ xs' ~ 9 ~D AS, ~~ Ill---:'; At, as ~ X5, ~ , S. As As' a. x S. ~ SCALE: 1:24 Yin AS,_ X5 X5 55, Xss 9 X,, X', '# . it, <' -- - ~L, my: 5,, 5 . o )~` X~ t~ --!C _ _ _ - ~' _ ! ~ Aloof a------' , ,_ _ _ _ , _, , . _ .,_-'----,'!_ ____' X,, ~ ~ Yes ~ 9 ~ Hiss cy d' K_ -~_~ ho ,.. ?. X~6 1, _ FIGURE 14 The west elevation of the Kin Ya'a tower kiva. The ruin of Kin Ya'a, originally built of native sandstone, lies within Chaco Canyon National Monument in New Mexico aIld probably dates to A.D. 100~1175. Variations in the wall surfaces are measured against a vertical datum plane, and edge details are carefully drawn in by the plotter operator. SOURCE: Koogle and Fouls Engineering, Inc., Albuquerque, New Mexico.
266 -t __ 1 . 3 o W[- . ~ o Cl . _ o ~ . ~ O , ~ O l . n . ; ~ l J ; ~ , n o o - ;, 14 [K~ . '_ 'z; ;< ,,~w '! ~ !, Z ,. .. ~ C: I , m < ~ _ ~ ~ U 9 U U U U ~ U U 't] ~ r~ ~_ .~. _ ~ b.ij ° C.) ~.= CO _ V ~ O P" C~ o o O ~ _ 3 o CO C~ - .~ - o ~ C) C) . ~ o a, ~ _ ,= O ~ v .= ._ C- V ~ a, ~ .= - 4 ._ ~ ~ o .° ~ U) CO ~ V ~ .- _ ._ :> ~ ~ ~. o ~ cn o o o.5 ~ a' ~i C~ cn _ ._ ~ C~ Z ~ ~ ~ ~ _ C' ~° V
Photogrammetr~c Measurement and Monitoring of Structures 267 Kiva at the Kin Ya'a site in New Mexico have been produced and will serve as the baseline data for future monitoring Figures 14 and 151. Both controlled line drawings and digitized point data derived by pho- togramrnetric means can be used in this method of monitoring struc- ture change. The frequency of monitoring these features by photogrammetric means is dependent on rate of change. To determine this, aerial and terrestrial photography will be reacquired (for example, every 6 to 12 months). Further, it is anticipated that planned mining activities in the Kin Ya'a area will have a negative impact, and during this activity the periodic rechecks wfl] be more frequent. It is obvious, of course, that effects of earthquakes, fires, or other disasters will be checked as soon as possible. One monitoring method that is practicable today is the use of elec- tronic distance measurement (EDM) equipment for the repetitive mea- surement of points on a structure. EDM devices make use of either a laser beam or microwave emitted from a transmitter, reflected from a prism or a point on the object to be measured, and then received at the transmission point. The time taken for the beam to reflect and return is accurately measured and, when atmospheric pressure, tem- perature, and humidity are corrected for, reveals the exact distance from the EDM to the target. Such equipment is used widely for sur- veying today, and typical accuracies are +0.5 cm in 1 km. Such devices could be used to measure the distance from a fixed station to a number of fixed points on a structure. Points to be measured would be located with respect to their importance in the integrity of the structure that is, points where a dimensional change would fore- tell significant alteration or collapse. Such points could be unobtru- sively provided with sockets into which a standardized prism would be fitted. A measurement would then be taken from the permanent instrument station (perhaps a socket into which the instrument would always be fitted for measurement) and the prism moved manually to the next point. While more field time would be required for such measurement than with aerial or terrestnal photogrammetry, far less laboratory time Photo developing, instrument setup and plotting, etc. ) would be expended. Many monitoring situations may not require the measurement of the potentially large number of three-dimensional point coordinates available in stereo photos, and EDM monitoring of a relatively small number of highly critical points on a structure or site could greatly speed the monitoring process. It would be feasible, for instance, to monitor a small number of points on a structure each day, or con-
268 CONSERVATION OF HISTORIC STONE BUILDINGS ceivably even several times a day, if this were deemed necessary. In addition, data collected with EDMs WOUi] be compatible with point data plotted photogramrnetrically. ZVlicrowave Scoping A microwave scanner, such as side-Iooking airborne radar or synthetic aperture radar, is essentially an EDM that scans the entire surface of a target area systematically and compiles a matrix of data on the distance from the scanner to the target. Although one commonly sees this matrix represented as a pseudophotographic product, it is recordable in digital form as well. If properly controlled, the digital output pro- duced by a ground-based microwave scanner could have certain ad- vantages over camera photogrammetry. A computer could produce almost instantly a comparison map of two digital point matrices re- corded from the same site or structure at different times; it is possible that pattern recognition programs could be developed that would reg- ister the two data matrices automatically with manual recourse to control points. Ground-based microwave scanners that would allow the scanning of specific structures with sufficient accuracy and resolution have not been developed, and a prototype program probably would cost millions of dollars. But this does not negate the possible use of microwave scanning to monitor structures and sites. If we are serious about de- veloping more efficient and effective methods for clearing with our historic heritage, a preliminary feasibility study might well be in order. Holography An intriguing process using laser technology for the measurement of physical objects, and the three-dimensional reproduction of their form, has appeared recently in the guise of holography. This technique entails focusing a split laser beam at an object from two directions; reflected light is then exposed on a photographic plate. Projecting a laser beam or other coherent light back through this plate, which contains no obvious image when viewed in normal light, reconstructs a "wave front" image of the object in three dimensions. The hologram can be viewed from different directions, and parts of the object that are hidden from one angle appear from others. Holography has been used to record historic artifacts at the Smith- sonian Institution and might offer a means of recording and monitoring historic structures as well. Before this could be accomplished, however, some startling jumps in laser technology would have to occur. To the
Photogrammetr~c Measurement and Monitoring of Structures 269 present, holograms have been made only of relatively small objects, the largest being humans. The only lasers with enough power to record larger things are those currently being experimented with by the mil- itary for purposes far removed from historic preservation. Another problem with holography for the measurement of large structures is that during the exposure of the hologram any movement within the scene (even on the order of the size of a wavelength of light) will disturb or ruin the plate. Ever-changing historic structures might be difficult to stabilize this perfectly for even an instant. Nonetheless, we predict that holography will become a useful aid to recording and monitoring structures and cultural resources in general. Reconstruction Digitized three-dimensional photogrammetric data can be used for more than simple representation of a structure or comparison with baseline data for monitoring purposes. In many cases the archeologist or cultural resource manager is concemed with structures that already have suffered considerable destruction or fabric change. Many sorts of theoretical statements are made on the basis of the forms of structures. These include population estimates, assumptions of the importance of communities in a trade network, and guesses about the degree of affluence or "refinement" of occupants. When making such assump- tions it is important that one know the original form of the structure, which is difficult to divine in certain instances. At Chaco Canyon National Monument in northwestern New Mexico, for instance, a matter of conjecture is the number of stories that pueblo ruins origi- nally had. An experiment under way at the Remote Sensing Division may help to solve this problem. Pueblo Alto Figures 12 and 13), will serve to illustrate this approach. As previously mentioned, the aerial imagery obtained served as the basis of photogrammetric mapping. During the course of this work, some 7,000 points were digitized in three dimensions and recorded on computer cards. Points along the tops of walls were digitized at each significant change in wall height, and other points were chosen within exposed rooms to establish the depth of rubble fill. Using only the wall-top points, a model of the ruin can be reconstructed by the com- puter. These wall-top points are being combined with data on rubble fill and information derived during test excavation in an attempt to re- construct the configuration of the original structure. While computer reconstruction cannot produce a single "true" picture of the original configuration of a pueblo, it can help suggest plausible alternative
270 ~ At_ a_ CONSERVATION OF HISTORIC STONE BUILDINGS m so.o ~ .o at - - Bow 1~° it ~1~° FIGURE 16 Computer perspective drawing of a small portion of the Pueblo Alto ruin. This drawing, constructed with a 10 x 10 mesh data resolution, shows little detail. SOURCE: Joe McCharen, Civil Engineering Research Facility, University of New Mexico. designs upon which theoretical statements can be based (Figures 16, 17, and 181. SUMMARY Both aerial and terrestrial photogrammetric techniques are invaluable for documenting historic and prehistoric masonry or other forms of architecture in plans and elevations with a range of detail and accuracy. As with any procedure, of course, there are limitations to the usefulness to ,` 0 - ye - .o ~ ~ l~ x~. FIGURE 17 A more detailed, 50 x 50 mesh resolution computer perspective drawing of the sense portion of Pueblo Alto appearing in Figure 16. Note that "hidden lines" are actually hidden in this rendition. Wall thickness and the irregular profiles of the partially ruined walls are apparent.
Photogrammetric Measurement and Monitoring of Structures 271 To _ r FIGURE 18 Employing digitized wall-top profiles such as those appearing in Figures 16 and 17, as well as room-fill data, the computer has reconstructed the four-room section of Pueblo Alto as it may have appeared when in use. VVEile this experiment is obviously on a small scale, reconstructions of entire prehistoric or historic structures could be postulated using similarly handled digitized photogrammetric information. Of available technological methods for assessing conditions and changes and for establishing baseline or quantifiable data upon which to for- mulate decisions for preservation efforts. Digitized photogrammetric point data in conjunction with computer programming gives the architect or archeologist a three-dimensional too] for analysis of structures. Digitized structural data also are useful for monitoring change or the effects of Herman or natural forces on fabric. Research in the application of microwave scanning, holography, and three-dimensional, computer-aided structural reconstruction also may offer the architect new tools for ensuring that the integrity of historic and prehistoric structures is protected. Finally, the methods described in this paper are nondestructive or ninimally hat to architectural materials. The techniques are at- tractive, of course, because one method of conservation especially nec- essary in archeology is the hands-off approach. REFERENCES 1. Borchers, P.E. 1977. Photogrammetric Recording of Cultural Resources. National Park Service: Washington, D.C. 2. Lyons, T.R., and T.E. Avery. 1977. Remote Sensing: A Handbook for archeologists and Cultural Resource Managers. National Park Service: Washington, D.C. 3. Wolf, P.R. 1974. Elements of Photogrammetry. McGraw-Hill: New York.
