British Columbia Specifications and Guidelines for Geomatics Digital Orthophoto

British Columbia Specifications and Guidelines for Geomatics Digital Orthophoto Content Series Volume 7 Digital Orthophoto Specifications Release 1.0 May 1997 1

Base Mapping and Geomatic Services Branch Ministry of Sustainable Resource Management CONTENTS 1.0 Introduction 2.0 Source Materials for Scanning 2.1 Acceptable source materials 3.0 Scanner Specifications 3.1 Geometric Accuracies 3.2 Radiometric Accuracies 3.3 Density Ranges 4.0 Photo Control Requirements REDO THIS PAGE 4.1 Control Transfers 4.2 Control Points 4.3 Soft Copy Derived Points 5.0 Scanning 5.1 Scanning Requirements 6.0 Image Orientations 6.1 Camera Calibration 6.2 Interior Orientation 6.3 Absolute Orientation 7.0 Image Rectification 7.1 Digital Elevation Model Requirements 7.2 Image Rectification Algorithm 7.3 Output Areas 7.4 Mosaicing 7.5 Quality Control 7.6 Orthophoto Accuracy 7.7 Pixel Sizes 2

8.0 Digital File Deliverables REDO THIS PAGE 8.1 File Formats 8.2 Media 8.3 File Naming 9.0 Hard Copy Output 9.1 Scales Allowed 9.2 Sheet Size 9.3 Plot Contents 9.4 Output Devices 9.5 Output Accuracy Appendix A Glossary of Terms Appendix B Sample ASCII Report File Appendix C User/Client Guidelines on Orthophoto/Mapping Projects Appendix D References 3

1.0 Introduction The following specifications cover the production of scanned aerial photos, digital orthophotos, and hard copy plots of digital orthophotos. The intention is to define the minimum requirements to meet the Ministry's needs and to maintain a consistently high level of quality for these products as produced for The Ministry of Sustainable Resource Management, Base Mapping and Geomatic Services Branch. These specifications shall prevail unless there is a written addenda to these specifications provided by the Ministry. Wherever brand or trade names are mentioned, these are meant for example only and are not to be construed as an endorsement of or being exclusive of other instruments or procedures that provide similar levels of quality. Whenever words or phrases are italicized, a definition of the term can be found in the glossary. 2.0 Source Materials for Scanning 2.1 Acceptable Source Materials Scanning shall be carried out only from first or second generation transparent originals. Scanning will be carried out from either an autododged original or the scan of the non-autododged original will be subsequently processed by autododging software. If the latter method is used, the dodging must be done in at least a 10 bit space (for each band if colour) in the scanner prior to outputting the final 8 bit files for storage. Scanning can be carried out on cut film duplicates in positive or negative form, or on original roll negatives or positives. If original materials are to be scanned then the scanner must not physically come in contact with the film while the film is in motion. The introduction of scratches to the original materials by the scanner will be unacceptable and cause for the rejection of the scans as well as possible penalties at the discretion of the Ministry. Scanning of existing hard copy orthophotos is not acceptable. Scanning from paper prints will not be allowed due to the reduced resolution of such materials. 3.0 Scanner Specifications (Refer to Specifications for Scanning Aerial Photographic Imagery) 3.1 Geometric Accuracies (Refer to Specifications for Scanning Aerial Photographic Imagery 3.2 Radiometric Accuracies (Refer to Specifications for Scanning Aerial Photographic Imagery) 3.3 Density Ranges (Refer to Specifications for Scanning Photographic Aerial Imagery) 4

4.0 Photo Control Requirements 4.1 Control Transfers One method of controlling aerial photography will be by means of transfer of existing photogrammetric control from an existing block of aerial triangulation (ie. TRIM). In no case shall the range of magnification exceed four times from one photo set to the other. Extreme care must be used when transferring control. A point transfer device approved by the Ministry for aerial triangulation must be used. A transfer quality designation must be made for each point. This can be in the form of an alphabetic code beside the point number on the print. Transfer codes will be as per Ministry AT specifications. Point marks on film for scanning shall have a size such that a minimum of a four by four raster array is used to cover the point mark in the original scan file. This is to ensure that the point marks will be retrievable for Quality Control assessment. An example would be that if the scan pixel size is to be 15 microns then the point mark must be at least 60 microns 4.2 Control Points There shall be a minimum of six points transferred to each frame for the purposes of orientation. It is preferred that four of these be inside, and near the corners of, the final sheet or frame if it is possible. The remaining points can be placed so as to provide a good three-dimensional resection of the image. A minimum of two points shall fall within the limits of each finished orthophoto sheet. 4.3 Soft Copy Derived Points If the aerial triangulation is carried out using pugless photogrammetric techniques there will be no need for physical marks on the ortho image for orientation purposes. However, in order to provide a means of quality control assessment, some physically measurable point is required on the orthophoto. In such a case where pugless AT techniques are used, the raw image scan shall have a five pixel wide by one pixel thick white (pixel value of 255) cross embedded in it prior to ortho rectification. The ticks shall be burned in at the plate coordinate of the points as derived from the AT data. Such a cross shall be burned into the image for a minimum of six points per frame as described in section 4.2 above. 5.0 Scanning (Refer to Specifications for Scanning Aerial Photographic Imagery 5.1 Scanning Requirements (Refer to Specifications for Scanning Aerial Photographic Imagery 6.0 Image Orientations 6.1 Camera Calibration When available, the camera calibration report for the camera in use shall be input to the system prior to orientations commencing. This shall include fiducial distances, radial distortions, and calibrated focal length. 5

6.2 Interior Orientation A minimum of four fiducials shall be measured on all frames. If fewer than four fiducials are visible, then all available fiducials will be measured. A note will be made of this deficiency and reported to the Ministry. The results of an affine transformation shall show an RMSE of less than 30 microns. A digital report shall be maintained and be available for delivery, if requested by the contracting authority. This report shall show the measured positions as well as the calculated fit to the calibrated positions. In the case where pugless AT results are to be used, all fiducials must be used to ensure a good tie to the original plate coordinate data as output from the AT process. 6.3 Absolute Orientation In the case of conventionally pugged aerial triangulation where the exterior orientation elements are available from the adjustment package a minimum of six points shall be read in each frame. These points shall be distributed as discussed in section 4. The results of the orientation shall show an RMSE of less than 20 microns at the photo scale. A digital report shall be maintained and be available for delivery, if requested by the Ministry. This report shall show the measured positions as well as the calculated fit to the control point positions. Parameters for earth curvature correction and refraction resulted from aerial triangulation shall be applied during the absolute orientation. While it is not required that an absolute orientation be completed on frames which are oriented using exterior orientation elements derived from the aerial triangulation adjustment, it is recommended that a check of the control points be done to verify that no error has been made in the selection of the photo or its orientation data. 7.0 Image Rectification 7.1 Digital Elevation Model Requirements In order to rectify an aerial photo and create a digital orthophoto a raster Digital Elevation Model (DEM) is required. This DEM shall meet the Ministry specifications for mapping for the final scale of the desired map. A DEM shall contain mass points and break lines sufficient to meet the forementioned specification. If a new DEM is to be compiled for the project, then it shall be compiled so as to meet Ministry specifications for the given scale and convert it to a raster DEM using appropriate scale and methodology similar to 1:20k raster DEM specifications. The DEM may also be derived from existing sources, if of a suitable nature. TRIM data shall be considered acceptable for the production of orthophotos at scales up to 1:10,000. All DEM's shall be collected on or converted to UTM NAD 83 prior to completing the digital orthophoto unless specifically requested otherwise by the Ministry. For areas where there are sheer vertical, raised, or sub-surface features, such as retaining walls, bridges, overpasses, underpasses, berms, etc. it will be necessary to adjust the DEM to ensure that the resulting orthophoto image appears graphically correct. The contractor shall be responsible for editing the DEM to ensure that such features appear visually correct. In such circumstances 6

the normal requirement for geometric accuracy will be made of lower importance than creating a graphically acceptable image. Under some circumstances the Ministry may prefer to have these features in their correct geometric position and forego the graphical qualities. In such a case the requirements for image quality with regard to image smearing will be relaxed. 7.2 Image Rectification Algorithm Commonly there are three resampling algorithms namely; nearest neighbour, bilinear and cubic convolution. Originally nearest neighbour was used as it was the simplest to program, and the fastest to execute without specialized hardware. However nearest neighbour results in serious image degradation under certain conditions. Bi-linear and cubic convolution methods were developed in part as a means to improve on the image quality and accuracy. Bi-linear was often used in the past as it took fewer computations and was thus faster than cubic convolution. However bi-linear produces a product that is half way between nearest neighbour and cubic convolution in terms of image quality. Based on practical experience and backed up by the majority of specifications from other government agencies throughout the world, that the cubic convolution is superior to the other methods. The improvement is specially dramatic on orthophotos at larger scales, or of linear features. Thus a typical TRIM II 1:20 000 sheet of hinterland areas may not be too noticeably affected, while the same scale orthophoto of an urban area would be more likely to show evidence of the problems. A large scale orthophoto of an urban area (i.e. greater than 1:10 000) would be even more likely to show the problem. As a result, image rectification shall be carried out using either cubic convolution, an equivalent, or better algorithm. A nearest neighbour resampling will not provide a good image quality output and is not acceptable. This is due to the image artifacts which are generally introduced into the resulting image by the nearest neighbour method. 7.3 Output Areas As most BCGS sheets are based on geographic boundaries, they are not rectangular in shape. Raster images however are, rectangular in nature. In such cases, the final raster image shall cover the minimum bounding rectangle of the BCGS sheet plus an over edge of 100 pixels. If the final sheet layout is to be based on nonstandard sheets where the sheets are on an angle, then the final orthophoto image shall fit the bounding rectangle with at least 100 pixels of over edge. The output orthophoto image rectangle shall, in this case, be parallel to the sheet edges and not orthogonal to the coordinate axis. 7

7.4 Mosaicing Where it is not possible to complete a full sheet from a single scanned image, multiple scanned images will have to be mosaiced together. Prior to completing a mosaic join the required images shall have their tones balanced, preferably by an autododging method that does differential tone and contrast adjustment throughout the images. The join line between the images in use shall be defined as a polyline. This line shall be chosen so as to minimise the obtrusiveness of the join itself. If feathering is used along the join line, it shall not result in any noticeable image degradation such as image blurring or double imagery. When a heavy forested area has been logged after the DEM compilation and before the orthophotography, then the mosaic lines shall avoid new roads in the cut blocks. In this way misalignment between frames caused by DEM inaccuracy will be minimized. 7.5 Quality Control All completed orthophoto images shall have the corrected positions of the ground control (pug points or "burned in" image crosses) verified against the calculated position from the AT process. A report shall be made on any points outside of the allowable tolerances for orthophoto accuracy in 7.6 below. This report shall be made available to the Ministry upon request. In addition to the above numeric check a visual inspection of the image shall be done. Areas of particular concern include imagery of inconsistent tone relative to its surroundings, and areas of apparently smeared or blurred imagery. Image smears are often caused by a spike in the DEM which must be fixed in onsultation with the Ministry. Smears can also be caused by extreme relief relative to the angle of view from the given photo. Large areas of such image smearing may be cause for rejection. In such a case an intermediated photograph, more closely centred on the smeared area, shall be rectified if available, and mosaiced into the final sheet. If no other image is available for repair of the smeared area then the Ministry may instruct the contractor to smooth the DEM in the affected area so as to reduce the extent or severity of the smearing. This will result in reduced planimetric accuracy in this area and is thus to be kept to an absolute minimum so as to not degrade the image accuracy. Adjacent images shall be checked to ensure that imagery is continuous and without gaps. Linear features within an image shall be checked for fit to existing map details, and if available (as in TRIM) features shall be checked to ensure that they tie within the allowed tolerance. Linear features such as roads, bridges, railways, overpasses, etc. shall appear straight and unwavering. In order for this to occur, the DEM shall normally include such features as break lines and may require editing prior to raster DEM creation and orthophoto rectification. 7.6 Orthophoto Accuracy The accuracy requirements in these specifications will reflect those standards set under the North American Treaty (NATO) Standard Agreement (STANAG) for the evaluation of Land Maps. Accuracies shall be measured only in the bivariate sense (please see TRIM specifications for more details on bivariate distributions). All orthophotos shall meet the following minimum accuracy standards. The Circular Map Accuracy Standard (CMAS) shall be equivalent to 0.5 mm at the final map scale. If these accuracy conditions cannot be met, this shall be reported to the Ministry prior to start of work on the project. 8

7.7 Pixel Size All orthophotos shall be produced with a pixel size in metres that fits the following criteria: Plot Scale 20,000 10,000 5,000 2,500 2,000 1,000 MaximumPixel 2.5 1.25 0.625 0.3125 0.25 0.125 Size metres metres metres metres metres metres In all cases the original scan resolution shall be at least as fine as the desired final Orthophoto pixel size less 20%. This specification is designed to allow for the production of high quality hard copies at the final scale. It also means that the image can be comfortably viewed at up to twice the planned scale in soft copy format. As an example, a 1:2,500 scale project would require an original scan pixel size of nominally 0.20 metres in order to achieve the 0.25 metre pixel size. If the scale of the source photography were 1:10,000, this would result in a minimum required scan resolution of 20 microns. Scanning at an even finer resolution is also acceptable. This requirement is made in order to ensure that an orthophoto pixel is created at a coarser resolution than the original scan. This will ensure that the desired pixel size is maintained even in areas of varying photo scales in the original photography caused by extreme relief. 8.0 Digital File Deliverables 8.1 File Formats All digital orthophotos shall be delivered in one or more of the following formats as selected by the Ministry at the time of contract award. Black and white images shall be delivered as 8 bits per pixel. Colour images shall be delivered as 24 bits per pixel (8 bits each for red, green, and blue). Multi-spectral imagery shall be delivered with at least 8 bits per band or by the dynamic range of the given image. Files shall be delivered as uncompressed TIFF files with GeoTIFF tags included for automatic image georeferencing. Arc Info files shall be delivered with an accompanying World file for automatic georeferencing of the image and several other packages. In addition, all images shall be accompanied by an ASCII report file indicating the coordinates of the four corners of the raster file, the pixel size in ground units, and the rotation matrix values used to orient the image to allow for georeferencing of the image in other image processing, CAD, or GIS programs. A sample of such a report is included in Appendix B. The Ministry may request that images be delivered in Intergraph COT format with a full set of overviews. Overviews are resampled versions of the full resolution pixels at ratios of 2,4,8,16 and 32 times the original resolution. This results in a final file that is 33% larger than the full resolution image. COT files shall contain headers with a Global Origin (GO) and Working Units (Units Of Resolution UOR) as specified by the Ministry. Other formats may be specified by the Ministry. 9

8.2 Media All files shall be delivered on one of the following media as selected by the Ministry at the time of contract award. CD-ROM ISO 9660 format 8mm Exabyte 2.3 Gb uncompressed format with the Tar command and blocking of 32 Kb (10Kb) 4mm DAT 2.1 Gb uncompressed format with the Tar command and blocking of 32 Kb (The above Tar formats are to be compatible with Unix and NT versions of the TAR software) 8.3 File Naming All digital orthophoto files shall be named according to their BCGS identification number. In order to properly accommodate the file names required for the sheets on conventional DOS-BASED systems the traditional DOS 8.3 character file naming convention be maintained. As a result of these criteria files are to be named with the following file name conventions. Examples: 1:250,000 BCGS sheet 82P.tif 1:100,000 BCGS sheet 82P_NW.tif 1:50,000 BCGS sheet 82P_1.tif 1:20,000 TRIM sheet 001.82P 1:10,000 BCGS sheet 0011.82P 1:5,000 BCGS sheet 00112.82P 1:2,500 BCGS sheet 001124.82P 1:2,000 BCGS sheet 001063.82P 1:1,000 BCGS sheet 0010632.82P If a special project is being produced that does not follow the BCGS sheet format then the file naming convention shall be established by the Ministry in consultation with the contractor at the start of the project. An example may be as follows: Project Specific sheet Telkwanw.tif 9.0 Hard Copy Output 9.1 Scales Allowed All hard copy output for the Ministry shall be at one of the standard BCGS scales as listed in section 8.3 above. The Ministry shall specify the desired scale at the time of request for proposals or on contract award, in consultation with the contractor. 9.2 Sheet Size All hard copy sheets produced shall be made to conform to the BCGS standard A1 sheet size. In the case of special projects the Ministry may specify a nonstandard sheet size at the RFP stage. 10

9.3 Plot Contents All hard copy plots shall, as a minimum, contain the following information: Date of Photography Scale of Photography Scale of hard copy plot UTM grid Zone Mapping Datum eg. North American Datum 1983 UTM grid at a 10 cm spacing at plot scale Annotation of the UTM grid on all four sides parallel to the sheet sides with an E or N appended to the end of the number Pixel size used to generate the plot Sheet or Area name or number Key map showing a three by three matrix of the current sheet and its surrounding sheet names Logo and name of the agency commissioning the mapping Optionally the Ministry may request the addition of the following layers of information: Contours Planimetric map information Cadastre Toponymy All line work plotted shall be plotted as per Ministry digital mapping specifications. All line work shall be plotted as a white line except for Cadastre which is to be plotted as a black line. The photographic image shall be masked so that it only fills the neat sheet area. Imagery beyond the BCGS sheet boundaries included in the digital orthophoto file, for example, will not be permitted. 9.4 Output Devices If the contract requires it, the plotting of the hard copy map sheet shall be done on a high resolution B+W film writer with a minimum of 2000 dpi of resolution. Imagery shall be plotted at a resolution of at least 200dpi. Line work shall be plotted at a resolution of at least 1500 dpi. Plotting shall be done at final scale, enlargement of a reduced original will not be accepted. Output is to be done to film negative form with film positives produced from the negative and, optionally, photographic paper prints, at the Ministry's discretion. If colour output is required, it is to be carried out on a plotter with a true continuous tone resolution of at least 200 ppi at the final scale. Colour plots on binary raster plotters such as the HP 750-C will not be acceptable as their true continuous tone resolution is in the order of 50 to 100 ppi. In the case of colour plots an opaque white paper or synthetic base is to be used to ensure high quality image detail is maintained. 11

9.5 Output Accuracy Final plots are to be produced such that they are dimensionally accurate to ± 0.1 mm throughout the image area for B+W plots. Film negatives and subsequent film positives shall be made on.18 mm thick base so as to ensure the dimensional stability of the plot. As there are currently no colour plotters capable of achieving such an accuracy level, the requirement for a colour plot shall be ± 3.0 mm throughout the plot area. 12

Appendix A Glossary of Terms Autododged Refers to a positive, negative, or digital file, that has had tonal and contrast variations throughout the image reduced by the effects of an autododging printer (such as the LogE), or has been processed in its digital form to the same effect. Bins A pixel bin is a given pixel intensity. It is usually used to refer to the number of occurrences of a given intensity. Burned A process whereby a graphic element such as a cross shape is digitally etched into a raster file. Density The photographic density of the transparency as measured by a densitometer. Density is equivalent to Where T= Transmissivity of the media. Feather A process used on the join between two images in a mosaic to help to reduce the difference in tone and contrast between the various images. First Generation Refers to a positive or negative transparency that has been produced from the original aerial negative or positive Geometric Accuracy Root Mean Square Error when a grid of known accuracy is scanned and the grid intersection positions in pixels are compared to the known position. Pixel A picture element, the digital representation of an image as an integer value. Normally within the range of 0 to 255 as stored in a single byte (8 bits). Mosaic A process whereby two or more images are joined into a single raster file in order to give the appearance of a single image. Pugless Procedure whereby photogrammetric control is generated using soft copy photogrammetric techniques. No physical "pug" points are needed or created. Raster A rectangular array of pixels representing an image by means of pixel intensities. 13

RMSE Root Mean Square Error Second Generation This refers to a positive or negative produced by contact exposure from the first generation original Step Wedge A standard photographic tool showing steps of increasing density. A typical wedge has steps at increments of 0.15 Density units within the range of 0 to 1.5 D. Transmissivity Percent of light transmitted by a transparent media 14

Appendix B Sample ASCII Report File -------------------------------------------------------------------------------- IMAGE REPORT FOR FILE: 0507.92H -------------------------------------------------------------------------------- Based on Design File Centroid of: 2143483.648 7443483.648 ORIGIN (Upper Left corner of Pixel): 702528.625 5487566.375 ORIGIN (Centre of Pixel): 702529.000 5487565.000 NUMBER OF ROWS : 16095 NUMBER OF COLUMNS : 20427 NUMBER OF PIXELS : 328772565 PIXEL SIZE : 0.75000000 SCANLINE ORIENTATION: Row-Major Upper Left Origin ROTATION (Degrees counter-clockwise positive X-axis): 0.000000000 X Y IMAGE EXTENTS UPPER LEFT (X Y) UPPER RIGHT (X Y) 702528.625 5487566.375 ----- Edge of Pixels ----- 717848.875 5487566.375 702529.000 5487566.000 ---- Centre of Pixels ---- 717848.500 5487566.000 LOWER LEFT (X Y) LOWER RIGHT (X Y) 702529.000 5475495.500 ----- Centre of Pixels ---- 717848.500 5475495.500 702528.625 5475495.125 ----- Edge of Pixels ----- 717848.875 5475495.125 TRANSFORMATION MATRIX c1 c2 c3 c4 r1 7.50000000000e-01 0.00000000000e+00 0.00000000000e+00-1.44095464800e+06 r2 0.00000000000e+00 7.50000000000e-01 0.00000000000e+00-1.95591764800e+06 r3 0.00000000000e+00 0.00000000000e+00 1.00000000000e-03 0.00000000000e+00 r4 0.00000000000e+00 0.00000000000e+00 0.00000000000e+00 1.00000000000e-03 15

ELEMENTS OF TRANSFORMATION MATRIX: r1c1 = pixel size * cosine of rotation angle r1c2 = pixel size * negative sine of rotation angle r1c3 = (not used for 2-dimensional image) r1c4 = X-coordinate of image origin r2c1 = pixel size * sine of rotation angle r2c2 = pixel size * cosine of rotation angle r2c3 = (not used for 2-dimensional image) r2c4 = Y-coordinate of image origin row3 = (not used for 2-dimensional image) row4 = (not used for 2-dimensional image) NOTE: * All coordinates are given in ground units *ARCINFO world files have the signs of r1c2 and r2c2 reversed (records 3 and 4 respectively in the world file) because the image Y-coordinate axis is positive downward in ARCINFO. 16

Appendix C User/Client Guidelines on Orthophoto/Mapping Projects As in any mapping project there are a number of factors that affect the final price of the project. Much as in the completion of DEM's, the cost is proportional to the number of photos, the accuracy required, the amount of detail required, the delivery format, the complexity of the terrain, etc. In order to assist the Ministry in effectively organising a project so as to assure the attaining of the project's goal while maintaining good cost control, the following concepts should be kept in mind. 1. Maximise the photo scale relative the final desired scale. A project will typically cost one quarter as much if the photo scale is doubled. With modern cameras and films it is reasonable to expect a top quality product with a magnification ratio up to eight times from the photo scale to the final scale for B+W photography. For colour we recommend the maximum magnification be kept below seven times. This recommendation is due to the lesser resolution of colour films. 2. Whenever possible design the sheet layout and the flight line layout so that a single photo can be used to produce a full sheet. This will remove the need for mosaicing and its extra costs. If it is not possible to complete a sheet from a single photo, endeavour to cover the sheet by not more than two or three photos. The larger the number of photos in the sheet, the higher the cost and the greater the chance that some mosaic lines will be noticeable. 3. Only plot the line work you need on the final ortho. The more complex the plotting required on the hard copy, the higher the cost will be. Remember too that having too much line work plotted on top of the ortho merely obscures the imagery and reduces its utility. 4. Whenever possible use existing aerial triangulation as a control base. This will remove the need for new aerial triangulation. Even if the AT data is a few years out of date, it is probably still sufficient for control of the orthophotos. 5. Whenever possible use an existing DEM for the production of the orthophoto. The DEM is usually one of the most expensive components of a mapping project so it is essential that it not be redone unnecessarily. If the DEM is meant only for the production of the orthophoto (that is no contouring or other DEM analysis will be done), then a less dense DEM can be used and thus costs reduced. 6. Each hard copy requested adds significant cost to the project. The least expensive option for hard copy is a single film positive directly produced on the plotter. The disadvantage is that if it is damaged and must be replaced it has to be replotted, which is expensive. If a negative exists, it is simpler and less costly to make a new film positive or, if multiple copies are required, a negative is often more cost effective. 17

7. On larger projects, where the economies of scale can come into effect, the best prices will be achieved. Smaller projects have inherently higher overheads relative to their size. An alternative is to group several small projects together to regain some of the economies of scale. 8. If no hard copy is required, do not order one as it is an expensive component to produce. Most agencies can make their own check quality plots on ink jet plotters as and when they need them. 9. When writing an RFP try to determine as precisely as possible the exact number of final sheets, the number of stereo models required, and the expected number of frames required for ortho production (usually using every other 60% overlap photo). If these numbers are not specified in the RFP different contractors will calculate based on differing estimates and you will not be getting as good a cost comparison. 10. Assign some weighting criteria for evaluating responses. Technical capability, experience, capacity, and methodology, should outweigh the importance of the price. 11. If a special project requires sheets laid out at differing angles relative the coordinate axis, try to lay out the photography to match. While some ortho production systems are limited to sheets orthogonal to the axis, most display and GIS systems are not. Take advantage of that flexibility. 18

Appendix D References Remote Sensing and Image Interpretation, Thomas M. Lillesand and Ralph Kiefer, Book published by John Wiley & Sons, Inc., 1995. Manual of Remote Sensing, Volume 1, Book published by American Society for Photogrammetry and Remote Sensing, 1996. Release Date: May, 1997 Last updated: March 06, 2002 19