Robert J. McGaughey

USDA Forest Service, Pacific Northwest Research Station

University of Washington, PO Box 352100

Seattle, WA 98195-2100



ABSTRACT: The appearance of landscapes and individual stands after harvest operations is critical to public acceptance of timber harvest practices. This paper reviews four visualization techniques suitable for visualizing the appearance of timber harvest operations: geometric modeling, video imaging, a hybrid technique combining geometric modeling and video imaging, and image draping. Techniques are reviewed for application to plot-, stand-, and landscape-scale project areas. Specific visualization software, both public domain and commercial applications, are identified. World wide web site addresses are given for most software packages.

Key Words: visualization, video imaging, timber harvest, visual simulation



The appearance of landscapes and individual stands after harvest operations is critical to public acceptance of timber harvest practices. Thorough planning, detailed site-specific analysis, and careful monitoring of harvest activities will not result in truly successful operations if the public views the resulting landscape as an eyesore. Attempts to mitigate the visual impact of harvests include modifying unit boundaries to conform to topography and other natural stand openings, prescribing silvicultural treatments that retain higher numbers of standing trees or groups of trees, and attempting to "hide" or "screen" harvest units from sight. These mitigation efforts can be successful. However, foresters charged with designing harvest unit shapes and silvicultural treatments often find it difficult to develop visually acceptable solutions by working in the field or with planimetric maps. The ability to visualize the appearance of a treatment or harvest operation before implementation can provide needed visual feedback during the design process.

People exhibit varying abilities to visualize the effect of harvest patterns and silvicultural treatments. Some individuals can look at maps and diagrams showing harvest unit boundaries and read descriptions of the silvicultural prescriptions to make a judgment of the appearance of the resulting operation but others require more picture-like visualizations. The proliferation of new technology such as geographic information systems and high-resolution satellite imagery have enabled forest managers to provide detailed maps and images showing a variety of resource characteristics. These traditional products, however, do not necessarily provide realistic representations of the appearance of forested areas before and after harvest operations. In addition, these products do not provide sufficient visual feedback to forest engineers and other resource specialists. Such feedback is essential if they are to design operations that produce a satisfactory visual condition.

Foresters, engineers, and other resource specialists can use computer visualization techniques to create images depicting the appearance of forested landscapes before and after harvest activities. Images depicting a variety of harvest unit designs help them preview the visual appearance of proposed activities. Visualization products, ranging from three-dimensional drawings to highly realistic rendered images, provide feedback while designing harvest units and silvicultural prescriptions. Such products also facilitate communication between the designer, other resource specialists, and the public.

This paper will present an overview of visualization techniques suitable for use in timber harvest planning, discuss their relative strengths and weaknesses, and provide guidelines for selecting an appropriate visualization technique given the terrain, stand, and operational conditions and the intended use of the visualization product.


Forestry professionals have used visualization techniques to address a variety of forest management problems. Prior to the advent of computerized methods, they used "artists' renditions" to communicate the effects of land management activities. Such visual and physical models, produced using artistic techniques such as perspective sketching, watercolors, air brush, and scale models, continue to communicate the spatial arrangement and extent of management activities to the lay public. However, the current trend in forest management is towards more detailed designs involving small treatment areas scattered over a larger landscape and the removal or modification of specific stand components. Alternative treatments utilize different mechanical methods, vary the spatial arrangement of treatment units, and specify different levels of modification within individual treatment units. With such treatments, the traditional "artists' rendition" cannot be made specific enough to represent the subtle differences between alternative treatments.

Computerized visualization methods range from simple diagrams to complete virtual realities. Four methods are suitable for producing visual representations of forest operations:

· geometric modeling,

· video imaging,

· geometric video imaging,

· image draping.

Geometric Modeling

Forestry professionals have used geometric modeling techniques to depict landscape conditions since 1976 (Myklestad and Wager, 1976). Geometric modeling methods build geometric models of individual components (ground surface, trees and other plants). The individual component models are then assembled to create a forest stand or landscape model. Scenes depicting the complete model are then rendered from a variety of viewpoints. In its simplest form, this technique can be used to generate perspective drawings showing typical GIS data coverages such as roads, streams, and polygon data overlaid onto the ground surface. More complex applications build very detailed models of individual trees that include small branches and leaves for use in rendering scenes or special effects for motion pictures. Geometric modeling systems specifically aimed at producing visualizations of forestry activities have been presented by several authors (Bonnicksen, 1993; Burk and Nguyen, 1992; Hanus, 1995; Kuehne, 1993; Fridley and others, 1991; Larson, 1994; McGaughey, 1997; McGaughey and Ager, 1996; McGaughey and Twito, 1988; Orland, 1997). These systems use perspective or orthographic drawing techniques to render stand and landscape images for land areas ranging in size from less than one acre to several thousand acres. Some systems are interfaced to stand projection models to show both the capabilities of the projection model and stand conditions resulting from modeled stand growth and silvicultural treatments over time. Some systems rely on commercial computer aided design programs such as Autocad to render images of the geometric models while others directly employ two- and three-dimensional drawing techniques to render images.

Video Imaging

Video imaging is a computer technique that uses computer programs to "cut-and-paste" or "paint" on scanned full-color video or photographic images to represent changes to the stand and landscape conditions. Video imaging produces television-quality (or better), full-color visual representations that depict current and future conditions. Orland (1988, 1993) reports the use of images created using video imaging techniques for both internal reviews of proposed management activities and for public information and participation.

Digital editing capabilities typically use a library of images representing different forest conditions to replace portions of an original image or to literally overlay objects or vegetation patterns onto an original image. To simulate the appearance of damage caused by defoliating insects, Orland and others (1990) used image processing techniques to analyze the color changes associated with insect damage in forests. They then applied similar color changes to new landscape images to simulate new damage. Larson and others (1988) report using similar techniques to simulate the effects of atmospheric pollution on the quality of photographic images and scenic views.

Geometric Video Imaging

A hybrid approach, called geometric video imaging in this paper, combines geometric modeling and video imaging techniques to produce very realistic images that accurately represent data describing the effects of forest management activities. Operators use geometric modeling techniques to produce scenes used to determine the location, arrangement, and scale of proposed landscape changes. Video imaging techniques are then used to modify a digitized image to reflect these changes. The technique can be extended to use geometric modeling techniques to determine the locations for digitized images, or icons, of single trees. Hybrid approaches result in images that accurately reflect the data describing proposed changes. However, to produce photo-like images, hybrid techniques require extensive libraries of tree images that represent an appropriate range of species, tree sizes, growth forms, and landscape positions. Orland (1997) describes the use of the SmartForest-II visualization system to guide editing efforts on a series of photo images used in public preference studies.

Image Draping

Image draping techniques mathematically "drape" an image over a digital terrain model and render the resulting scene from a variety of viewpoints. Operators usually derive the image from a satellite scene, scanned aerial photograph, orthophoto, or scanned map sheet. Several GIS and image processing applications provide draping capabilities. Most include rectification procedures to properly orient and align the digital image to the ground surface. Simple applications utilize orthophoto images that have already been registered to the ground surface and corrected for elevation, or relief, displacement. Bishop and Flaherty (1991) report the development of an image draping technique that relies on a library of textures to provide the image content needed to create realistic representations of GIS databases.

Image draping techniques can produce visualizations suitable for depicting landscape-scale vegetation patterns. Operators simulate treatment effects by modifying the original image using techniques common to video imaging to reflect harvest activities. It is difficult to model plot- and stand-scale treatment effects using image draping techniques. Details like differences between individual trees or groups of trees cannot be shown because textures are simply mapped onto the ground surface.

Comparison of Visualization Techniques

Table 1 compares the data requirements, operational complexity, level of realism, and data integrity of various visualization techniques. If a technique has high data requirements, this means that either large amounts of data or very detailed data are needed to apply the technique. Realism ratings refer to the image quality relative to a photograph of a similar scene. Operational complexity represents a combination of general software system complexity, data manipulation required to us the technique, and general artistic abilities required of the operator. Data integrity refers to the technique's ability to represent changes in the source data describing a particular treatment in the final image. High data integrity means that the technique can represent small changes in the input data.

Direct comparison of visualization techniques is difficult. Techniques based on some type of geometric modeling generally rely on detailed data describing terrain and stand conditions to create an image representing those data. Video imaging and image draping techniques start with an image and attempt to modify the image to correctly represent proposed changes. All of the techniques mentioned are suitable for forestry visualization. However, inherent limitations in the techniques, data requirements or quality of the final image products make some techniques better suited to particular applications in forestry.

Table 1. Characteristics of visualization techniques suitable for depicting the effects of forest harvesting operations.

Visualization technique Data requirements Level of realism in final scene Operational complexity Data integrity
Geometric modeling High Low to moderate1 Moderate to high High
Video imaging Low High Moderate Low to moderate2
Geometric video imaging Moderate to high High Moderate to high Moderate to high
Image draping Low to moderate Moderate Moderate to high Moderate

1 Commercial rendering applications used in motion picture special effects can produce very realistic scenes. However, their high cost and the complexity of the data needed to build the required models limits their usefulness in forestry applications.

2 Video imaging techniques rely heavily on the operator's skill to manually make changes in the image to represent changes in stand or landscape conditions. A skilled operator can produce visualizations that accurately represent proposed conditions. However, there are no internal mechanisms inherent in the technique to ensure that modified images accurately represent proposed treatments.


Operators cannot apply geometric modeling applications embedded into or that rely on a forest growth model to a broad range of visualization projects. Growth models represent specific forest types and geographic regions and cannot be easily applied to other areas or forest types. Applications that rely on commercial CAD packages may be difficult for some individuals to apply. In most cases, the operator is assumed to have access to and be proficient with the CAD application. Operators not familiar with the CAD application must learn to use its features and the procedures to convert the forest or landscape data into a properly formatted geometric model.

Video imaging techniques, usually applied to landscape-scale projects, can be used to represent stand treatments. However, it is difficult to represent small scale, subtle changes affecting individual trees and small groups of trees. Video imaging relies heavily on the operator's skill to visualize the effect of a proposed treatment and make the necessary changes to an image to accurately reflect the treatment. Operators must be familiar with the operations and forest conditions before they can make image modifications that accurately and consistently reflect changes in the underlying data. Simply painting a new texture over a portion of a stand may not adequately communicate changes in stand attributes such as tree spacing and crown closure.

Video imaging techniques lack accurate methods for transferring planimetric information such as treatment unit locations and roads onto the perspective scene represented by the original photograph. Most often, operators manually transfer such features from topographic maps. This manual process can lead to significant errors. The consequences of such errors can be as simple as a misplaced harvest unit. However, incorrectly placed treatment unit boundaries can have a significant impact on the appearance of the unit from specific viewpoints. For example, the position and shape of the unit boundary significantly affects the portion of the unit that can be seen from a roadside viewpoint. Positional errors can result in final images that significantly misrepresent the visual effect of a proposed treatment.

Image draping techniques show unit locations and overall vegetation patterns. Because image draping techniques apply image texture to the ground surface only, they represent very little height detail for trees. For example, the edge of a clearcut usually provides a sharp contrast between the height of an adjacent stand and the bare ground of the clearcut area. Image draping techniques cannot display this height difference. Image draping is also prone to errors when attempting to model the effect of a light source shining onto the terrain surface. Aerial images used to color the ground surface already represent the effect of a light source: the sun. Additional attempts to further shade the terrain surface based on another light source or a different sun angle can create optically confusing scenes. Simply using the intensity of the aerial image will result in a scene with some lighting effects. However, images will often be of low contrast and appear "flat" because aerial photographs are often taken during mid-day to minimize shadows. This lighting problem also makes it difficult to modify an aerial image to reflect the effect of treatments. Areas of an image used to provide textures for a treatment area must have the same orientation or aspect as areas being replaced in the original image. The lighting effects and conflicting shadows in final image will be confusing to viewers if the orientations are not similar.

By combining geometric modeling and video imaging techniques, geometric video imaging can produce data-driven images that exhibit a high degree of realism. A skilled operator is still needed to modify digital images. Large libraries of images are needed to provide the rich palette of vegetation types, sizes, orientations, and colors needed to make final image modifications realistic. Nonetheless, this technique is the only visualization techniques currently available that can produce photographic quality images that reflect small changes to stand and landscape data.

All four methods presented in this paper can use databases describing stand conditions before and after proposed changes to provide a data-driven solution. The degree to which changes in the database are visible in the final image depends on the technique and the operator's skill at applying the technique. Systems must have well-designed databases and linkages between the database and the visualization technique. Such systems allow users to respond quickly to design changes and shifts in management strategies and to provide consistent results for a variety of treatment alternatives. In general, geometric modeling techniques, including geometric video imaging, are considered the most "data-driven". They provide for a one-to-one relationship between data describing a treatment and objects in the final scene. Video imaging and image draping techniques do not provide such a relationship but instead rely on the skill of the operator to accurately reflect database elements in the final scene.

Visualization Project considerations

Significant criteria to be considered when selecting an appropriate visualization technique for a project are the:

· size of the project area,

· overall goal of the visualization products,

· amount of detail that must be present in the final visualizations,

· amount of data available describing the project area.

Table 2 summarizes these criteria for three project scales: landscape, stand, and plot. The land areas covered by these scales are loosely defined. Projects can span more than one scale and the same data set can be used to generate images representing different scales. For example, many projects include landscape-scale images to show the overall vegetation patterns and harvest unit locations and stand- or plot-scale images to show harvest unit layout information and specific stand treatments.

As a general rule, the larger the project area, the less detail required in the input data and the final visualizations. Landscape-scale projects usually show the spatial arrangement, scheduling, and cutting intensity of treatment areas. Such projects can be accomplished using geometric modeling techniques based on digital terrain data and stand descriptions consisting of average tree size and stem density. The same project could be accomplished with little or no descriptive data using video imaging techniques. The operator would simply modify photographs of the project area to show the location of harvest units and to reflect the effect of the treatments. Additional photographs showing treatments similar to those being considered provide the image content used to edit the original photographs. Stand-scale projects, on the other hand, require more detailed descriptions of stand conditions. Tree size, species composition, and possibly spatial arrangement are needed to represent the effects of harvest activities on overall stand structure. With such detail, foresters can use the images to make judgments regarding the habitat quality for a particular species or other site-specific interpretations. Projects designed to show detailed changes to stand structure, for example, small areas to be thinned adjacent to large, highly desirable crop trees, require an even more data describing stand and tree characteristics.

Table 2. Characteristics of visualization projects representing different land areas.

Project scale Land area Overall goal Tree/plant detail Typical data requirements
Landscape > 200 ha

> 494 acres

Vegetation texture, spatial arrangement of stand types, location of specific treatment areas, visual quality, insect or other stand damage effects Species, height, color, density Topography; ground surface characteristics; stand polygons; average tree size, predominant species, and stem density for each stand
Stand 2-200 ha

5-494 acres

Harvest area layout, patch clearcut or group selection treatments Species, height, color, density, crown characteristics Topography; ground surface characteristics; stand polygons; tree size and species distributions for each stand, general understory conditions
Plot < 2 ha

< 5 acres

Stand structure, habitat quality, silvicultural prescriptions Species, dbh, height, crown characteristics, foliage characteristics Individual tree characteristics, individual or aggregated understory characteristics, spatial arrangement of understory and overstory plants


To some extent, the intended use of images produced by a visualization project dictates the technique used to produce the images. Geometric modeling techniques are sufficient to communicate the intent and specific details of harvest operations and silvicultural treatments. Such images work well for internal reviews involving resource specialists and others familiar with forest practices. Different types of images may be needed for public presentation and review. Such uses require images that are more realistic to engage the viewers and provide them with enough information to evaluate management alternatives. The lay public may have difficulty relating the somewhat abstract images produced using geometric modeling techniques to their own, in-woods, experiences. Most people readily understand images that closely resemble photographs. However, the use of highly realistic images can lead to misconceptions of the amount of control foresters have over the future conditions of stands and landscapes. It is often difficult to convince a lay reviewer that the project area will not look exactly like a photographic image created using video imaging techniques. Their expectations may, in turn, far exceed what is physically and biologically possible for the project area.

Visualization software

Many software packages are available that can produce forestry visualizations. Commercial computer aided design (CAD), rendering, and animation systems produce and render geometric models to create images and animation sequences. Unfortunately, commercial systems are expensive, often required a specialized operator to produce satisfactory results, and require extensive data manipulation to convert typical forestry data into a usable geometric model. Public domain systems provide visualization and image editing capabilities suitable to forestry visualization and are usually available for little or no cost. Table 3 provides a summary and contact information for several visualization tools.

For video imaging applications, Adobe Photoshop is the most common software used for image editing and manipulation. The Gnu image manipulation program (GIMP) for UNIX platforms provides many of the capabilities found in Photoshop and is available free of charge. Additional image manipulation programs such as Adobe PhotoDeluxe, Corel PhotoPaint, and Softkey Photofinish are available from a variety of vendors but none provide the full range of editing capabilities found in Photoshop.

Many GIS and image processing applications provide visualization capabilities. Geographic information systems such as ARC-INFO provide to ability to create perspective views showing the ground surface and vector and raster data layers. Image processing systems such as ERDAS and IDRISI provide image rectification and draping capabilities. However, GIS and image processing systems typically cannot render objects such as trees on the ground surface.

Conclusions and recommendations

This paper has reviewed many techniques suitable for producing visualizations depicting harvest operations and other forest management activities. Visualization techniques and software systems are rapidly evolving as personal computers become more powerful. Recent computer developments such as three-dimensional rendering and image processing functions included as standard features with inexpensive display adapters are making previously impossible levels of realism and rendering speed commonplace. Even with the most sophisticated visualization systems, the amount of agreement between projected conditions, represented by visualizations, and attainable conditions can vary dramatically. Once a desired visual condition has been identified, achieving the desired condition can be difficult given the operational constraints imposed by forestry equipment, vegetation response to the treatment, topography, and operator proficiency. Harvest systems have specific requirements to ensure safe, efficient operations. When the desired visual condition requires very specific harvest activities and patterns, operations can become unprofitable or, in the worst case, dangerous for logging crews.

Table 4 summarizes the authorís recommendations regarding appropriate visualization techniques given the size of the project area, amount of data available for a project, and amount of realism required in the final images. Video imaging can be used for any project for which data describing terrain and vegetation characteristics is not available. For stand- and plot-scale projects, it may be difficult to accurately reflect the effects of silvicultural treatments using video imaging. When data is available, geometric modeling or geometric video imaging techniques should be used to provide the most data-driven visualizations. Image draping techniques are best reserved for projects designed to show an overview of a large project area with few details regarding the treatment or treatment effects.

Computer visualization techniques can be an extremely powerful tool to communicate and educate critics of forest operations. However, they can just as easily be used to mislead people into believing a harvest operation will have little or no detrimental impact on the appearance of a forested landscape. Practitioners must ensure that visualizations are accurate representations of reality. This does not mean that visualizations must exhibit a high degree of realism to be effective. Images must, however, accurately represent the best data available describing stand and landscape conditions and the effect of a harvest operation on these conditions.


Table 3. Software packages for forestry visualization.

Software package Visualization technique Scale Computer platform Cost1 Additional information
Stand visualization system (SVS) Geometric modeling Plot PC-DOS Free
UTOOLS and UVIEW Geometric modeling Stand or landscape PC-DOS Free

SmartForest Geometric modeling Stand or landscape UNIX (SGI or IBM-RS6000 with OpenGL) Free
Landscape management system (LMS)2 Geometric modeling All scales PC-Windows Free
Adobe Photoshop Video imaging All scales PC-Windows, macintosh, UNIX $$
Gnu Image Manipulation Program (GIMP) Video imaging All scales UNIX Free
Paint Shop Pro Video imaging All scales PC-Windows Free, $$
USFS, Southern Research Station visualization system Geometric modeling Stand or landscape UNIX Free
VistaPro3 Geometric modeling and image draping Landscape PC-DOS, PC-Windows, macintosh $$
IDRISI Image draping and perspective views Landscape PC-DOS, PC-Windows $$
Persistence of vision raytracer (POV-Ray) Geometric modeling All scales Many platforms Free

POV-Ray is a general purpose ray-tracing system capable of producing detailed, realistic images of geometric models.

VisualFX Geometric modeling Stand or landscape PC-DOS $$ Available from author: John Heasley (303) 223-3149
CLRview Geometric modeling Stand or landscape Silicon Graphics IRIX Free
Visual Explorer Image draping and geometric modeling Landscape PC-Windows Free, $$
TruFlite Image draping Landscape PC-Windows Free, $$

1 System cost refers to the purchase price of the software. Free packages are either public domain or otherwise freely available. Software marked with "$$" are commercial products available at retail outlets or from the software producer. Items marked with both "free" and $$ indicate that the product is available as a free trial version as well as a commercial version.

2 LMS uses SVS to provide plot-scale visualizations and UTOOLS/UVIEW to provide stand- and landscape-scale visualizations.

3 VistaPro does not provide for individual tree placement or specification of individual tree characteristics making it difficult to use to accurately representing a variety of stand conditions.


Table 4. Recommended visualization techniques given project scale, amount of data available to describe the project area and proposed activities, and the amount of realism required in the final images.

Scale Amount of data Required realism Recommended technique
Landscape Low Low to high Video imaging, or image draping
Landscape High Low Geometric modeling
Landscape High High Geometric video imaging
Stand Low Low to high Video imaging
Stand High Low Geometric modeling
Stand High High Geometric video imaging
Plot Low Low to high No suitable techniques
Plot High Low Geometric modeling
Plot High High Geometric modeling or geometric video imaging


Literature Cited

Bishop, I.D; Flaherty, E. 1991. Using video imagery as texture maps for model driven visual simulation. In: Buhyoff, G.J., ed. Resource Technology 90, Proceedings of the Second International Symposium on Advanced Technology in Natural Resources Management; 1990 November 12-15. Washington, DC. Bethesda, MD: American Society for Photogrammetry and Remote Sensing. 58-67.

Bonnicksen, Thomas M. ©1993. Restoring ancient giant sequoia forests. An electronic publication available from Thomas M. Bonnicksen, Department of Forest Science, Horticulture/Forest Science Building, Texas A & M University, College Station, TX 77843-2135.

Burk, Thomas E.; Nguyen, Man V. 1992. Visualizing the operation of a distance-dependent tree growth model. The Compiler. 10(2): 10-19.

Fridley, James L.; McGaughey, Robert J.; Lee, Frank E. 1991. Visualizing engineering design alternatives on forest landscapes. American Society of Agricultural Engineers paper #917523. Presented at the 1991 International Winter Meeting. Chicago, IL. December 17-20. 7 p.

Hanus, Mark L. 1995. Generation of an animation interface for ORGANON. Corvallis, OR: Oregon State University. M.S. thesis.

Kuehne, M.J. 1993. High quality forestry: an alternative for management of national forest lands in western oregon and western washington. Available from: Northwest Independent Forest Manufacturers, PO Box 11346, Tacoma, WA 98411.

Larsen, David R. 1994. Adaptable stand dynamics model integrating site-specific growth for innovative silvicultural prescriptions. Forest Ecology and Management. 69:245-257.

Larson, S.M; Cass, G.R; Hussey, K.J; Luce, F. 1988. Verification of image processing based visibility models. Environmental Science and Technology. 22(6):629-637.

McGaughey, Robert J. 1997. Visualizing forest stand dynamics using the stand visualization system. In: Proceedings of the 1997 ACSM/ASPRS Annual Convention and Exposition; April 7-10, 1997. Seattle, WA. Bethesda, MD: American Society for Photogrammetry and Remote Sensing. Vol. 4:248-257.

McGaughey, Robert J.; Twito, Roger H. 1988. VISUAL and SLOPE: perspective and quantitative representation of digital terrain models. Gen. Tech. Rep. PNW-GTR-214. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. 26 p.

McGaughey, Robert J.; Ager, Alan A. 1996. UTOOLS and UVIEW: analysis and visualization software. In: Proceedings of the sixth biennial USDA Forest Service remote sensing application conference; April 29-May 3, 1996; Denver, CO. Bethesda, MD: American Society of Photogrammetry and Remote Sensing. 319-329.

Myklestad, E.; Wager, J.A. 1976. PREVIEW: computer assistance for visual management of forested landscapes. Res. Pap. NE-355. Morgantown, WV: U.S. Department of Agriculture, Forest Service, Northeastern Forest Experiment Stations. 12p.

Orland, Brian. 1988. Video-imaging: a powerful tool for visualization and analysis. Landscape Architecture. 78: 78-88.

Orland, Brian. 1993. Synthetic landscapes: a review of video-imaging applications in environmental perception research, planning, and design. In: Marans, Robert W. and Daniel Stokols, eds. Environmental simulation: research and policy issues. New York.: Plenum Press. 213-252.

Orland, Brian. 1997. Forest visual modeling for planners and managers. In: Proceedings of the 1997 ACSM/ASPRS Annual Convention and Exposition; April 7-10, 1997. Seattle, WA. Bethesda, MD: American Society for Photogrammetry and Remote Sensing. Vol. 4:193-203.

Orland, Brian; Daniel, Terry C.; LaFontaine, Jeanine; Goldberg, C. 1990. Visual effects of insect damage in western mixed coniferous forests. Final Report. Cooperative Research Project, USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, Research Work Unit 4501. Imaging Systems Laboratory, Department of Landscape Architecture, University of Illinois, Urbana. 33 p.

Virtual Reality Laboratories, Inc. ©1993. Vistapro user guide (IBM DOS Version 3.0). San Luis Obispo, CA: Virtual Reality Laboratories, Inc. [paged by individual chapters].