H. M. KARARA Department of Civil Engineering

University of Illinois at Urbana-Champaign Urbana, IL 61801

Close-Range Photog rammetry: Where Are We and Where Are We Heading?" Great strides have been made in close-range photogrammetry in the past few years, and continued vigorous progress is fully anticipated in this field.

P AOTOGRAMMETRY is much more than just a method for the production of topographic maps

of the Earth's surface; it is a general physical mea- suring system with applicability to the most diverse fields of science, technology, and art.

In close-range photogrammetry, the range of the object-to-camera distance is limited. Some advocate 300 metres as a lnaximuln limit, while the minimum distance is essentially zero (say a fraction of a mil- limetre) to encompass macro- and microscopic pho- tographs.

and expensive, and were not ideal for close-range photogrammetry.

Analog plotters currently used in close-range pho- togrammetry can be grouped in three categories (Karara, 1979): (1) normal-case plotters, which are often a part of a close-range photograminetric camerdplotter system, such as the Wild A40 ster- eoplotterlCl20 stereometric camera system; (2) uni- versal plotters, such as the Zeiss C-8 Stereoplani- graph, which have fairly large ranges of tilt, depth of field, principal distance, and base setting, and which are used for aerial, terrestrial, and close- range photogrammetry; and (3) topographic or map-

ABSTRACT: During the past two decades, close-range photograinrnetry has under- gone an evolution in its very systems, and vigorous, continued progress is expected in this field. This paper discusses the development of analog methods and instru- ments and the evolution of the analytical approaches. The gradual acceptance of non-metric cameras as data acquisition systems in close-range photogrammetry is discussed, analyzed, and evaluated. The most recent techniques used to improve the precision and reliability of analytical solutions are summarized. The emerging trends in close-range photogrammetric systems are presented and briefly discussed.

Since the 1920's, analog methods were ~aramount in close-range photogrammetry. universal plotters, which were primarily designed for aerial photog- raphy, were made so general in their design to en- able their use in terrestrial and close-range photo- grammetry. Because of the universality of their de- sign, these instruments were rather complicated

* Based on a paper published in the Close-Range Pho- tograinmetry and Surtjeying Workshop Handl?ook, 1984 ASP-ACSM Fall Convention, San Antonio, Texas.

PHOTOCRAMMETRIC ENGINEERING AND REMOTE SENSING, Vol. 51, No. 5, May 1985, pp. 537-544

ping plotters, designed for aerial mapping, accept vertical or near vertical photography, and can be used for close-range photogrammetry, such as the Jenoptik Topocart stereoplotter. About the same time that the normal-case stereoplotters were de- veloped, stereornetric cameras with a fixed base and a fixed focus (such as the Zeiss SMK-40 stereometric camera) were introduced. Such camerdplotter sys- tems played a major role in the development of close-range photogrammetry in architecture, ar- chaeology, biostereometrics, civil engineering, in- dustry, foresnic science, and numerous other areas of application in the period 1920 to 1960.

0099-11 12/85/5105-0537$02.25/0 63 1985 American Society for Photogrammetry

and Remote Sensing

PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING, 1985

Electronic computers started a new phase of de- velopment in photogrammetry. Stereoplotters soon began to be equipped with model-coordinate re- cording devices, and numerical computations of co- ordinate transformation, areas and volumes, etc., were easily accomplished. Such computations are undertaken either off-line in computer-assisted ste- reoplotters or on-line in digital stereoplotters with fully analytical reconstruction of stereo-models and digital plotting tables with numerically controlled line drawing and symbol generation, such as the Kern DSR-11 analytical plotter.

Unlike analog stereoplotters, analytical stereo- plotters have no limitation on model size, plot size, overlap, and base setting, and can be used to plot nearly any type of close-range imagery, including non-metric photography. Software development for close-range applications has been increasing at a fast rate during the past few years, and computer pack- ages andlor computational services are becoming in- creasingly available from equipment manufacturers as wellas from consulting firms.

The analytical reconstruction of the bundle of rays and of the stereomodels generated from image co- ordinate readings in comparators andlor digital plot- ters opened new possibilities for close-range pho- togrammetry. Any type of camera with any orien- tation can be used. Sensor systems are not limited to camera and film, but X-rays, scanning and trans- mission electron microscopes, and solid state cam- eras are also suitable for photogrammetric measure- ments.

Digital methods have opened new possibilities for the integration of photogrammetric and other com- putations, leading to a more effective measuring process in relation to the purpose of the total pro- cedure. A case in point here is computer program GEBAT, developed at the University of New Bruns- wick and the National Research Council of Canada (El-Hakim, 1982), computer program SAPGO (Wong, 1974), and computer program GENTRI (Larsson, 1983). Optimized geometry of photography and use of the multi-approach (multi stations-more than the two stations necessary to generate a stereo- model, multiframe photography, multi measure- ments, multi models, etc.), as recommended by Hottier (1976) and Torleghd (1981a, 1984), and dig- ital filtering for gross error detection are measures that can be taken to improve the precision and re- liability of the results.

As indicated by the author (Karara, 1979), the analytical approach can be used in the simple case where interior and exterior orientation elements are

Metric as well as non-metric cameras are used to acquire photography in close-range photogram- metry. According to Faig (1975a), a non-metric camera has an interior orientation that is completely or partially unknown and which is frequently un- stable. It should be noted that the traditional defi- nition of a metric camera as havine fixed and con- " stant interior orientation no longer holds because there are currently focusable cameras that definitely must be regarded as metric cameras, such as the Wild P31 and Jenoptik UMK 1318, among others.

A major characteristic of metric cameras is that the radial distortion is so small that it often may be neglected for most practical applications (Torelgird, 1981b). For high-accuracy requirements, several so- phisticated models have been developed to account for radial and tangential lens distortions and as well as film deformations. Details on this topic are avail- able in the Manual of Photogrammetry (Slama, 1981), Brown (1982), Faig (1975), El-Hakiin and Faig (1977), and others.

Early metric close-range cameras had fixed focus and, thus, their range of applications was rather lim- ited. Only during the last decade has it become pos- sible to produce focusable metric cameras.

A listing of single and stereometric metric cam- eras available on the market in 1979, together with their major characteristics, is given in the Handbook of Non-Topographic Photogranzmetry (Karara, 1979).

Most available metric cameras use glass plates be- cause photographic plates have been traditionally preferred over film as a recording medium for close- range photogrammetry on account of their superior lateral dimensional stability. Recent investigations by Gates et al. (1982) and Brown (1984a) refuted this myth. For ultra-precise work using microflat plates of 0.25 inches thickness over a 9 by 9-inch format, departures from the best fitting plane in ex- cess of k0.0005 inch are common (Brown, 1984a). By measuring the "topography" of the photographic surface, the plane is, in effect, mathematically flat- tened to virtual perfection. In his newest close- range film camera CRC-1, Brown (1982, 1984b) in- corporated an ultra flat reseau platen. This feature, in effect, renders the film as dimensionally stable as plates. The flatness of the film in the CRC-1 camera at the moment of exposure is reported to be within k 0.0001 inch, and thus far superior to even inicro- flat plates (Geodetic Services, Inc., 1983).

known, but its advantages and versatility become much more pronounced in the most general case of NON-METRIC CAMERAS

photogrammetry in which a simultaneous solution Flexibility in focussing, high resolution, and rel- incorporates the interior and exterior orientation atively low prices are the main features which led elements of all photographs and the space coordi- to the ever-increasing use of amateur (or non- nates of object points, all as unknowns. metric) cameras in non-topographic photogram-

CLOSE-RANGE PHOTOGRAMMETRY

metry. The recent significant advances in analytical data reduction schemes have helped non-metric cameras become viable data acquisition system in close-range photogrammetry. It has been demon- strated, both theoretically and experimentally, that non-metric cameras can yield moderately high rel- ative accuracies (in the range of 1:6,000 to 1:10,000 of the photographic distance or better) when used in conjunction with suitable analytical reduction schemes of increasing sophistication. With the use of the "multiple approach" (Hottier, 1976; Torle- gird, 1981, 1984), these accuracies can be improved significantly.

Among the algorithms developed for data reduc- tion of non-metric photography are the Direct Linear Transformation (DLT) developed at the Uni- versity of Illinois (Abdel-Aziz and Karara, 1971), the 11-parameter solution (Bopp and Krauss, 1978a, 1978b), and a simplified version of the 11 parameter solution (Adams, 1981). Bundle adjustment solu- tions originally developed for analytical aerotrian- gulation are also used for data reduction from non- metric photography (Faig, 1975; Moniwa, 1977; El- Hakim, 1982; Gruen, 1984; among others). As Adams and Riither (1984) aptly put it, "The elegant bundle solution implies a good knowledge of the camera constants and is really not suited to the simple non-metric camera whereas the fully con- trolled projective transformation methods makes the non-metric camera a useful measuring tool."

Non-metric cameras have been used extensively in biostereometrics (e. g., Karara and Herron, 1974; Coblentz and Herron, 1978; Herron, 1983; Karara, 1974), in mining engineering (e.g., Murai et al., 1980; Brandow et al., 1976), in traffic accident in- vestigations (e.g., Waldhausl and Kager, 1984), in animal husbandry (e. g., Hatzopolous, 1982), and in numerous other fields.

Software for data processing from non-metric 1 photography has been developed for a number of

analytical plotters such as the MAC0 35/70 (Gillen and McGlone, 1984) and Kern DSR-1 (Fuchs and Leberl, 1984) and has resulted in increasing the appeal and the areas of applications of non-metric cameras.

As Faig (1975) aptly put it, "calibration provides the link between metric and non-metric cameras." Calibration of a camera encompasses the computa- tion of the parameters of interior orientation. The basic parameters of interior orientation of a camera are the principal point and principal distance (cali- brated focal length or camera constant), as well as radial (symmetric) and decentering (frequently con- sidered in form of its asymmetric and tangential components) lens distortions, and film deformation. In addition, &nity and non-perpendicularity of the comparator axes used for image coordinate mea- surement are included for the system calibration.

In contrast to aerial cameras, for which laboratory and field calibration procedures have been stan- dardized over the years, no standard calibration pro- cedures exist for close-range cameras. Calibration of close-range cameras is usually carried out in one of three fashions: in the laboratory, on the job, or by self calibration.

Laboratory calibration is completely separated from object photography, and is ideally suited for metric cameras focused (or focusable) at infinity. Besides goniometers, collimator banks, and similar arrangements, test areas of various sophistication have been used for laboratory calibration. (e.g., Abdel-Aziz and Karara (1974), Karara and Faig (1972), Faig (1971), and TorlegPrd (1967)). The mathematical formulation is normally based on the collinearity equations, with each object-space con- trol point providing two equations. A minimum of five object-space control points is required to solve for the basic parameters of interior orientation (prin- cipal point and principal distance). With the inclu- sion of additional parameters of the interior orien- tation, the number of object-space control points has to increase accordingly (Linkwitz, 1972).

On-the-job calibration utilizes photography taken of the object and the object-space control sirnulta- neously. As least one full (X, Y, Z) object-space con- trol point should be provided for every two un- known quantities included in the solution. The mathematical formulation is essentially the same as for laboratory calibration. On-the job calibration is used for metric as well as non-metric cameras. More information is given by Faig (1976).

Self calibration differs significantly from the pre- vious ones in that it does not require object-space control as such for calibration. Multiple (at least three) convergent photography is taken of the ob- ject. Using the collinearity condition equations and well defined object points, elements of the interior orientation are computed for the camera used (metric or non-metric). This method is well docu- mented in the literature (e.g., Brown (1971, 1972), Kolbl (1972, 1974), and Faig (1975b).)

Self and on-the-job calibration determine the in- terior orientation of the same photographs which are used for object measurement. This eliminates the effect of instability of non-metric cameras. These types of calibration may be combined with a partial calibration for radial lens distortion, thus reducing the number of unknown interior orientation param- eters and the number of observations necessary for a strong solution (Torlegird, 1981b).

Non-conventional images are produced by im- aging systems which do not use a lens and an image plane and thus are not frame photographs based on the central projection of the object onto an image plane (Torlegird, 1981b). To this group belong X-

rays, scanning and transmission electron micro- scopes (SEM and TEM), television systems, digital images, continuous strip photography, and holo- graphic and 1noir6 techniques.

A representative sample of projects in industrial photogrammetry are briefly discussed here to illus- trate the use of close-range photogrammetry in the industrial and engineering fields. In the first two projects non-metric cameras were used to gather photogrammetric data; the third, fourth, and fifth projects utilized a commercially available camera; while the sixth and seventh projects resorted to cus- tomer-built large format metric cameras in order to meet ultra high accuracy requirements.

This project was connected to an investigation by the Illinois State Geological Survey on geologic data and structures that could be used to predict un- stable conditions in coal-mine roofs. Sites at which roof falls have already occurred were studied to ob- tain evidence that might point to such predictors. Traditional measurements with the Brunton com- pass can be overwhelming, especially when the sites to be measured are hazardous. In this project, strike and dip of planes were determined photogram- metrically using a 70-min non-metric Yashica C camera with an 80-mm focal length and j73.5 lens. Four to ten targets were used to define each plane of interest. Object-space control was obtained by using a modular aluminum frame consisting of eight cubic sections that could be assembled in various configurations. Photo-coordinates were measured in a Wild STK-1 stereocomparator and reduced by the Direct Linear Transformation (DLT) method (Abdel- Aziz and Karara, 1971), which involves 11 transfor- mation unknowns for each photograph. The space angle between the orientations of the planes, as de- termined on the basis of photogrammetric method and the Brunton compass, ranged from lo to 3O and indicated that an analytical photogrammetric system utilizing a small-format non-metric camera meets the accuracy requirements of geologic work, and that this method is not only technically feasible but in fact practical to use in geological mapping. Full information about this project is given by Brandow et al. (1976).

A Kodak bellow-type press camera of 200- by 250- mm format equipped with an Ektar lens having a 241.3-mm focal length was used to measure the three-dimensional movements of jointed rocks sur- rounding a tunnel model. This measurement pro- gram was in support of a research study concerning

movement behavior of jointed rocks surrounding an underground tunnel. The camera was mounted 3.84-m above the tunnel model, and stereoscopic pairs of vertical photographs were taken. The B + H ratio was 0.15 and the average photo scale was 1/16. The three-dimensional position of about 350 targets on the tunnel model were determined from a stereoscopic pair of photographs for each loading. Comparison of the computed positions of successive loadings then provided a measure of the amount and pattern of displacement of each target point.

Thirty-one control points whose X, Y, and Z co- ordinates were computed (estimated standard de- viation in X and Y was + 0.6 mm; relative elevations were estimated to an accuracy of 20 .5 lnln at one standard error). In addition, 15 check points were also located in the stereoscopic coverage area. A Wild STK-1 stereocomparator was used for coordi- nate measurements. Computer program SAPGO (Wong, 1974) was used for data reduction. The er- rors in computed motion vectors at the 15 check points with sample size of 75 (six stereopairs were taken under different loading conditions resulting in five riiotion vectors per check point). The true values of these vectors were zero. The results were as follows (mean error * standard deviation): X: -0.09 2 2.4 mm, Y: + 0.6 r 1.8 mm, Z: + 3 2

9 mm. This accuracy was considered completely ac- ceptable.

Interesting results were obtained when the pho- togrammetric measurements were compared with dial gauge measurements. The gauges were located near the checkpoints and had a least count of 0.0025 mm. A comparison of the two independent sets of measurements showed that the gauge readings were consistently too large by about 3.5 mm. After the gauge measurements had been corrected for con- stant bias, the photogrammetric measurements were in excellent agreement with the gauge read- ings. The differences compiled for 25 movement vectors has a mean of -0.001 mm and a standard deviation of 20 .5 mm.

PRED~CTING THE FIT OF SHIPS BUILT IN HALVES

The ease and cost with which the two halves of a ship may be joined depends largely on the dimen- sional equivalence of the mating faces. Advance de- tailed knowledge of this equivalence before the halves are launched and brought together allows the builder to undertake adjustments andlor to plan specific joining procedures to accommodate known mismatches. In this project, mating faces of two 126,000 DWT tankers, both of which were built in halves, were photographed using a Wild P-31 Uni- versal Terrestrial Camera. Convergent photography was taken from eight stations around the ship, four from ground stations and four from a personnel car- rier slung from the boom of a "cat" crane. The eight

CLOSE-RANGE PHOTOGRAMMETRY

photographs were measured on a monocomparator. A bundle adjustment program was used. These sur- veys provided a wealth of as-built dimensions of each face with a tolerance of 3/32 inch (2.4 mm). Data for mating faces were compared by means of computer processing to produce digital and graph- ical displays of the predicted fits. More about this project is given by Kenefick and Peel (1978).

In a similar project, the midship section was di- mensioned. In order to increase the payload of ex- isting ships, it has become commonplace to cut a ship in half (fore and aft sections), insert a midship section, and then weld the three units together. This project was undertaken to reasonably assure that the midsection would be dimensionally equiv- alent to the mother ship. Here again, the Wild P- 31 Universal Terrestrial Camera was used to take eight photographs from four ground stations and four over-head stations of the midship section. A bundle adjustment program was used. A resultant RMS of the photographic measurement residuals of 1.8 micrometres was obtained after processing. The typical accuracy of a triangulated target was 0.03 inches (0.76 mm) (standard deviation) in the vertical plane and 0.08 inches (2.03 mm) to and from the observer. Refer to Kenefick (1977).

USE OF PHOTOGRA.MMETRY IN THE MANUFACTURE OF

HIGH PERFORMANCE AIRCRAm

McDonnell Aircraft Company is using photo- grammetry in the periodic, recycling (re-inspection) of aircraft assembly tools. A Wild P-31 camera is used to acquire convergent photography. A Wild Aviolyt BC1 analytical steroplotter complete with CRABS (Close-Range Analytical Bundle Solution) software (developed by J. E Kenefick) is used for data reduction. The system is currently being used on a production basis in connection with the man- ufacture of the F-15 Eagle, the T-18 Hornet, and the AV-B8 Harrier I1 aircrafts which require large, specially designed assembly tools. Accuracy (one sigma standard deviation) produced by the bundle adjustment is converted into tolerances by a mul- tiplier (2.5). The average tolerances achieved has been: X = 20.004 inches, Y = 2 0.007 inches, and Z = t0.003 inches. The Y-axis is in the direction of the depth of the tool. These tolerances fall well within the expected limits and also satisfy tool re- cycle inspection measurements (Powell, 1984).

MEASURING COMMUNICATION ANTENNAS

A custom-built 9 by 9 inch (23-cm by 23-cm) glass plate terrestrial camera was used to photograph the dish in three different orientations: looking straight up and at elevation angles of 60" and 40". Most of the photographs were taken from a helicopter. Three highly convergent photographs were exposed for each orientation of the dish. A bundle adjust-

ment was used. Accuracies achieved averaged about 1 part of 120,000 of the diameter (D) of the dish. If needed, accuracies in excess of D/200,000 can be obtained by increasing redundancy in observations using the "multi-concept" as discussed by Hottier (1976) and Torlegird (1981a). For more information, refer to Geodetic Services, Inc. (1983).

MONITORING OF GROUND SUBSIDENCE OVER

UNDERGROUND EXCAVATIONS

Three-dimensional vectors of ground subsidence over underground excavations, typically mines and salt domes, were determined for the U.S. Bureau of Mines. Accuracies ranging from D/180,000 to Dl 250,000 (D is the maximum dimension of the object) were produced from sets of 16 or more photographs taken by a custom-made 9 by %inch (23-cm by 23- cm) glass plate camera with 480-mm focal length. The camera was helicopter-borne. On each such project, from 20 to as many as 500 targets were distributed over the area of interest, which typically covers several hundred acres and reaches out to stable ground. Each area was photographed at least twice to improve the project's reliability. For more information, refer to Geodetic Services, Inc. (1983).

EMERGING TRENDS IN CLOSE-RANGE PHOTOGRAMh4ETRY

Studying the recent progress in the various as- pects of close-range photogrammetry, the following trends emerge:

Continuation of the efforts to improve precision and reliability of close-range photogrammetric methods; Development of modular fully integrated data re- duction systems to meet modest and moderate, as well as the ultra high accuracy, requirements; Continued work to improve off-line and on-line data reduction systems for both metric and non- metric photography; Continuation of the work on development and im- provement of software packages for close-range photogrammetry for various analytical plotters, and for metric and non-metric photography; Improvement and development of software pack- ages for ~nicrocomputers (~nicrocomputer tech- nology is advancing at an extremely rapid rate and it is anticipated that, as software is developed, many new applications will evolve); Development of software packages for data reduc- tion of close-range non-conventional imageries; Integration of photogrammetry into production sys- tems; Improvement and development of digital on-line photogrammetry and robotics; Continuing the development and improvement of systems for underwater photogrammetry; Study of ways and means of obtaining results more rapidly; Study of cost factors and presentation of proposals to management; and Improvement of co~nmunications between photo-

PHOTOCRAMMETRIC ENGINEERING & REMOTE SENSING, 1985

grammetrists on the one hand and engineers and scientists on the other hand in the various disci- plines using close-range photogrammetry.

Perhaps a rather definite indication of the direc- tions to be taken by close-range photogrammetry in the next few years can be obtained from the reso- lutions adopted by Commission V (Non-Topographic Photogrammetry) of the International Society for Photogrammetry and Remote Sensing (ISPRS) at the XV ISPRS Congress held in Rio de Janeiro, Brazil in June 1984. Among the resolutions pertinent to the topic of this paper are:

Resolution T. Vll: ANALYTICS OF NON-TOPOGRAPHICAL

PHOTOGRAMMETRY. Recommendation: Maintenance of a working group to continue the exploitation of mathematical developments and new data reduction schemes in non-topographical photogrammetry, giving special emphasis to prediction and assess- ment of reliability and precision, design of optimal photogrammetric systems taking technical and eco- nomic aspects into account, and expanding the studies of on-line point positioning.

Resolution T. Vl2: REAL TIME AND DIGITAL CLOSE-

RANGE PHOTOGRAMMETRY. Recommendation: Real time aspects of digital photogrammetric processing should be given high priority in all relevant activi- ties organized by Commission V, especially in the monitoring and control of processes in scientific, in- dustrial, and biomedical applications.

Resolution T . V/3: APPLICATIONS OF CLOSE-RANGE

PHOTOGRAMMETRY. Recommendation: Establishment and maintenance inter alia of activities in the fields of biostereometrics, architecture, quality control in industrial production, and the monitoring of struc- tures.

Resolution T . V/4: MARKETING STRATEGY FOR CLOSE-

RANGE PHOTOGRAMMETRY. Recommendation: Com- mission V should disseminate information on the advantages of successfully developed new systems, for commercial development, and to provide evi- dence for potential users (for example the medical profession, professional architects, civil engineers, production engineers, etc.) to demand the instru- mentation required, by presenting creditable ac- counts of new developments in the meetings of ap- propriate learned societies.

Considering the recent exciting activities and emerging trends, it is difficult not to predict that the future of close-range photogrammetry is bright indeed.

Abdel-Aziz, Y. A., and H. M. Karara, 1971. Direct Linear Transformation from Comparator Coordinates into Object-Space Coordinates. Symposium Close-Range Photogrammetry. American Society of Photogram- metry, Falls Church, Va., pp. 1-18.

, 1974. Photograminetric Potentials of Non-Metric Caineras. Civil Engineering Studies, Photogram-

metry Series No. 36, University of Illinois, Urbana, Il., 120 p.

Adams, L. P., 1981. The Use of Non-Metric Cameras in Short Range Photogrammetry. Photogrammetria, 36:51-60.

Adams, L. P., and H. Riither, 1984. Two Phase Photo- grammetry with Displaced Control. International Ar- chives of Photograminetry and Remote Sensing, 25(A5):19-28.

Bopp, H., and H. Krauss, 1978a. An Orientation and Cal- ibration Method for Non-Topographic Applications. Photogrammetric Engineering and Remote Sensing, 44(9):1191-1196.

, 1978b. Extension of the 11-Parameter Solution for On-the-Job Calibrations o f Non-Metric Cameras. International Archives of Photogrammetry, 22(5):7-12.

Brandow, V. D., H. M. Karara, H. H . Damberger, and H . E Krausse, 1976. A non-Metric Close-Range Pho- togrammetric System for Mapping Geologic Structures in Mines. Photograinmetric Engineering and Remote Sensing, 42(5):637-648.

Brown, D. C., 1971. Close-Range Camera Calibration. Pho- togrammetric Engineering, 37(8):855-866.

, 1972. Calibration of Close-Range Cameras. Inter- national Archives of Photogrammetry, 19(5):Unbound paper:26 p.

, 1982. STARS (Simultaneous Triangulation and Re- section System), A Turnkey System for Close-Range Photogra~nmetry. International Archioes of Photograin- metry, 24(5/1):68-89.

, 1984a. Unflah~ess of Plates as a Source of System- atic Error in Close Range Photogrammetry. Interna- tional Archives of Photo-grammetry and Remote Sensing, E(A5):Unbound paper:29 p.

, 1984b. A Large Format, Microprocessor Controlled Film Camera Optimized for Industrial Photogram- metry. International Archilies of Photograininetry and Remote Sensing, E(A5):Unbound paper: 29 p.

Coblentz, A. M . , and R. E. Herron (Editors), 1978. NATO Symposium on Applications of Human Bio- stereoietrics. SPIE Proceedings, Volume No. 166. So- ciety of Photo-Optical Instrumentation Engineers, Bel- lingham, Washington, 374 p.

El-H&n, S . E , and W. Faig, 1977. Compensation of Sys- tematic Image Errors Using Spherical Harmonics. Pro- ceedings ASP Fall Technical Meeting, pp. 492-499.

El-Hakiln, S . F., 1982. The General Bundle Adjustment Triangulation (GEBAT) System-Theory and Applica- tions. NRC 20539, National Research Council, Ottawa, Canada, 47 p.

Faig, W., 1971. Calibration of Close-Range Photogram- metric Cameras. Syinposium of ASP-Uniwrsihj of Illi- nois on Close Range Photogrammetry, Urbana, Illinois, pp. 111-131.

, 1974. Precision Plotting o f Non-Metric Photog- raphy. Proceedings of lSP Corninission V Symposium Biostereometrics '74, Washington D.C., pp. 54-64.

, 1975a. Photogrammetric Equipment Systems with Non-Metric Cameras. ASP Symposium on Close-Range Photograinmetric Systems, Champaign, Illinois, pp. 648-657.

CLOSE-RANGE PHOTOGRAMMETRY

, 1975b. Calibration of Close-Range Photogram- metric Systems-Mathematical Formulation. Photo- grainwwtric Engineering and Reinote Sensing, 41(12): 1479-1486.

Fraser, C. S., 1984a. Multiple Exposures in Non-Metric Camera Applications. lnternational Archives of Photo- graininetry and Remote Sensing, 25(A5):286-295.

, 198413. Network Design Optimization in Non-To- pographic Photogrammetry. lnternational Archives of Photogrammetry and Remote Sensing, 25(A5):296-307.

Fuchs, H., and L. Leberl, 1984. Universal Analytical Plotter Software for Photographs with Perspective Geometry (CRISP). Proceedings, 50th ASP Annual Meeting: pp. 774-78 1.

I Gates, H. W. C., S. Oldfield, C . Forno, P. J . Scott, and S . A. Kyle, 1982. Factors Defining Precision in Close- Range Photogrammetry. International Archives of Pho- togrammetry, 24(5/1):185-195.

Gates, J. W. C., and K. B. Atkinson, 1984. Photogram- metry in Science and Technology: a Report on the Ac- tivities of ISPRS Commission V (1980-84). lnternational Archities of Photogrammetry and Reinote Sensing, 25(A5):315-324.

Geodetic Senrices, Inc., 1983. Services, Software and S ys- terns for Industrial Photogrammetry Offered by Geo- detic Services, Inc., Melbourne, Florida, 15 p.

1 Gillen, L G. and McGlone, J. C.. 1984. Operation of the ?vIACO 35/70 Small Format Analytical Stereoplotting System. Proceedings, 50th ASP Annual Technical Meeting, pp. 332-341.

Gruen, A., 1984. Processing of Amateur Photographs. In- ternational Archities of Photograinmetry and Remote Sensing, S(A5):Unbound paper:14 p.

Hatzopolous, J . N., 1982. Mapping Exotic Animals by Close-Range Photogrammetry. Proceedings, 48th ASP Annual Meeting: pp. 28-34.

Herron, R. E. (Editor), 1983. Biostereoinetrics '82. SPIE Proceedings #361. Society of Photo-Optical Instru- mentation Engineers, Bellingham, Washington, 359 p.

Hottier, P., 1976. Accuracy of Close-Range Analytical Res- titution: Practical Experiments and Prediction. Pho- togrammetric Engineering and Remote Sensing, 42(3): 345-375.

Karara, H. M., 1974. Aortic Heart Valve Geometry. Pho- togrammetric Engineering, 40(12):1393-1402.

, 1975. Industrial Photogrammetry. ASP Sym- posium on Close-Range Photograminetric Systems, Champaign, Il., pp. 97-141.

(Editor), 1979. Handbook of Non-Topographic Photogrammetry. American Society o f Photogram- rnetry, Falls Church, Va., 206 p.

Karara, H. M., and W. Faig, 1972. interior Orientation in Close-Range Photogrammetry: An Analysis of Al- ternative Approaches. lnternational Archives of Pho- togrammetry, 19 (Part 5, Report W G V/1):23-33.

, 1980. An Expos6 on Photographic Data Acquisi- tion Systems in Close-Range Photogrammetry. Inter- national Archioes of Photograminet ry, 23(B5):402- 418.

Karara, H. M. , and R. E. Herron, (Editors), 1974. ISP Coinmission V Symposium Biostereometrics '74. American Society o f Photogrammetry, Falls Church, Va., 642 p.

Kenefick, J . E, 1977. Application of Photogramrnetry in Shipbuilding. Photograminetric Engineering and Re- mote Sensing, 43(9):1169-1175.

Kenefick, J . F., and D. D. Peel, 1978. Predicting the Fit of Ships Built in Halves. lnternationul Archives of Photogrammetry, 22(V), Unbound paper, 25 p.

Kolbl, 0 . 1972. Selbstkalibrierung von Aufnahmekam- mern. Bildmesstrng und Luftbildwesen, 40(1):31-37.

, 1974. Tangential and Asymmetric Lens Distor- tions, Determined by Self Calibration. Proceedings of ISP Coinmission 111 Symposium, Stuttgart.

Larsson, R., 1983. Siinultaneotts Photograininetric and Geodetic Adjustlnent - Algorithms and Data Struc- tures. Dissertation, Department of Photogrammetry, Royal Institute o f Technology, Stockholm, Sweden, 99 p.

Linkwitz, K., 1972. Some Remarks on Present Investi- gations on Calibration o f Close-Range Cameras. In- ternational Archives of Photogrammetnj, 19 (Part 5, Report o f W G V/1):53-55.

Moniwa, H., 1977 Analytical Photogramtnetric Systems with Self-Calibration and its Applications. Ph.D. Dis- sertation. University o f New Brunswick, Fredericton, N . B., Canada, 127 p.

Murai, S., H. Nakamura, and Y. Suzuki, 1980. Analytical Orientation for Non-Metric Camera in the Applica- tions to Terrestrial Photogrammetry. lnternational Ar- chives of Photograminety, 23(B5):515-525.

Murai, S., R. Matsuoka, and T. Okuda, 1984. A Study on Analytical Calibration for Non-Metric Camera and Ac- curacy o f Three Dimensional Measurement. Znter- national Archities of Photograinmetry and Reinote Sensing, 25(A5):570-579.

Powell, G. E., 1984. The Use of Photogrammetry in the Manufacture o f High Performance Aircraft. Proceed- ings, 50th ASP Annual Meeting, pp. 655-659.

Slarna, C. S. (Editor), 1980. Manual of Photogrammetry. American Society of Photogrammetry, Falls Church, Va., 1056 p.

Torlegiird, K., 1967. On the Determination of Interior Orientation of Close-Up Cameras Under Operational Conditions Using Three Dimensional Test Objects. Fotograinmetriska Meddelanden, 5(3), 100 p.

, 1981a. Accuracy lmproveinent in. Close-Range Photograminetry. Schriftenreihe HSBw Vermessung, Miinich, Heft 5, 68 pp.

, 1981b. Development of Non-Topographic Pho- togrammetry and its Future. Finnish Society of Pho- togrammetry, 50th Anniversary Publication, pp. 49- 75.

, 1984. Multi-models to increase accuracy - Sum- mary o f an experiment. lnternational Archives of Pho- tograininetry and Remote Sensing, 25(A3b):1055- 1064.

Waldhausl, P., and H. Kager, 1984. Metric Restitution o f Traffic Accident Series from Non-Metric Photographs. lnternational Archives of Photogrammetry and Re- inote Sensing, 25(A5):732-739.

Wolf, P. R., and S . A. Loorner, 1975. Calibration of Non- Metric Cameras. Proceedings of ASP Symposium on Close-Range Photograminetric Systems, Champaign, Il., pp. 373-393.

PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING, 1985

Wong, K. W., 1974. SAPGO-A Computer Progra~n for chives of Photograminetry and Remote Sensing, the Simultaneous Adjustment of Photogrammetric and W(A5):779-788. Geodetic Observations. Civil Engineering Studies, wong, K. w., and A. p. Vonderohe, 1978. Measuring of Photogrammetry Series No. 38, UILU - ENG - 74 - Displacements Around Tunnel Models by Motion Par- 2003, University of Illinois at Urbana-Champaign, Ur- allax. international Archives of Photogrammetry, bana, 11. 138 p. 22(5): 12-21.

, 1984. Photogrammetric Measurement of Move- (Received 26 July 1984; accepted 6 January 1985; revised ments in Tunnel Model Testing. lnternational Ar- 25 January 1985)

New Sustaining Member

Williams-Stackhouse Incorporated

2118 Mannix Drive, San Antonio, TX; (512) 824-6301

W ILLIAMS-STACKHOUSE INCORPORATED is a multi-disciplined mapping firm which offers a wide range of surveying and photogrammetric services to both government and private industry.

The company was founded in 1938 under the name V. L. Beavers Engineers, and was initially a general surveying and civil engineering firm. In the early 1950s the company expanded its capabilities into the field of photogrammetry, and in 1960 assumed its present name.

WSI's staff consists of approximately 30 full-time employees including photogrammetrists, surveyors and geodesists, computer technicians, cartographers, and photo lab technicians.

WSI's offices and production facilities are located in a company owned, 10,000 square foot building adjacent to San Antonio International Airport.

For aerial photography WSI uses the Wild RC-10 and Wild RC-8 camera systems, with 3W, 6", and 12" focal lengths. Also, the company is adept at using color, false color, and color multispectral films as well as black and white.

The company maintains a complete in-house photo lab which includes such equipment as a LogEtronic Mark IV contact printer, a LogEtronic strip printer, a K&E HE-12 enlarger, a Kargl rectifying enlarger, two DuPont automatic film processors, and a custom-built, multi-lens copy camera.

For field control, WSI is equipped with Hewlett Packard 3820A Total Stations for horizontal surveys, and Zeiss Ni-2 and Wild N-2 levels for vertical surveys. The company is also experienced in the use of satellite positioning and inertial guidance surveying systems, and has employed them successfully on a number of projects.

For analytical aerotriangulation, the company utilizes both semi-analytical and fully-analytical programs on a VAX 111780 computer system.

For photogrammetric compilation, WSI is equipped with five stereoplotters (Kern and Zeiss) each of which is encoded for digitizing.

WSI operates throughout the United States with most of its work being in the southern part of the country. WSI also works outside the U.S. and is presently performing several projects in Latin America and one in Africa.

WSI has long been regarded as a people-oriented, client-oriented company. Its goal is to remain at the leading edge of surveying and photogrammetric technology, while at the same time offering practical, economical solutions to its clients' needs.