Joint Education Initiative

home *** CD-ROM | disk | FTP | other *** search

/ Joint Education Initiative / Joint Education Initiative.iso / dos / idrisi / idrisi.txt < prev next >

Wrap

Text File | 1990-06-20 | 67KB | 1,394 lines

MODULE DESCRIPTIONS (JEDI Disk Subset) Each of the IDRISI modules in the JEDI Disk Subset set is described in the following section. Modules are arranged in alphabetic order. Characters contained in quotes are intended to indicate the literal characters that one would type in response to a user prompt. The quotes should themselves be omitted. Thus, for example, "km" in- dicates that one should type km, not quote-km-quote. Error messages that are particular to a module are listed. In cases where the message is self-explanatory, no further explanation is given. If you encounter an error message that is not listed, it is a general IDRISI message. Refer to the ERROR MESSAGES section for these. COLOR Version : 3.xx (Different versions exist for various display adaptors. During installation, the version chosen (eg. COLOREVA.EXE) is normally renamed to be COLOR.EXE. The filename of each version, as it is distributed, is listed with each of the specific device descriptions below.) Purpose : COLOR is a display module intended for the rapid examination of images in color on selected hardware devices. Titles and legends stored in the documentation files of images will also be displayed. The spatial and color resolution will vary considerably, depending upon the particular device being used. Because of these differences, separate module descriptions are given for each version. Note that the EGA/VGA driver is the preferred version of COLOR for working with IDRISI. It offers substantially more features as well as higher resolution. Command Line Parameters : Color normally expects to display an image of integer color codes. For example, the version of COLOR for the IBM EGA display adaptor can display 16 colors, and therefore expects an image composed of integer codes within a range of 0 - 15. STRETCH can be used to generate such an image. However, COLOR can also produce an immediate image of any file through autoscaling. By specifying an "a" command line parameter after its name (ie., "COLOR a") autoscaling will be invoked. With autoscaling, the minimum and maximum values in the documentation file for the image are used to classify all image values (either integer or real) to an appropriate color. This is good for a "quick look" but generally STRETCH produces a more informative result. EGA/VGA VERSION : COLOREVA.EXE Version 3.06 Operation : This version of COLOR works with both EGA and VGA graphics adaptors. It should be renamed COLOR.EXE for use (this is done automatically in the installation process). This module can automatically sense whether an EGA or VGA adaptor is present and adapt accordingly. COLOR begins by asking for the name of the image to be displayed. It then asks for the palette to be used. If an EGA card is in effect, the default IDRISI palette, the standard IBM palette and a user-defined palette will be offered. If a VGA is present an additional grey scale palette will also be offered. If the user-defined palette is chosen, you will then be asked to enter the name of the palette file (see the notes section below concerning its format). COLOR next asks if you wish a legend and the expansion factor to be used in displaying the image. For the expansion, enter an integer number (eg., 2, 3, 4 etc). If the image is too large to fit into the available display space, you will then be prompted to select the start and end columns and rows for the displayed image. After the image is displayed, the normal text screen can be restored by pressing return. The EGA/VGA version of COLOR also supports a wide range of operations while the image is being viewed, such as on-screen digitizing, image inquiry, and vector file overlay. The on- screen digitizing and interactive inquiry capabilities both require a Microsoft-compatible mouse. See the notes section below for specific details. The interactive query feature is important for identifying feature positions when using RESAMPLE. The on-screen digitizing feature is almost essential for defining training areas when using the supervised classification features of the Image Processing Ring. It can also be used very effectively for updating existing images as well as digitizing features that are visible on remotely sensed images. Notes : 1. Pressing the letter "c" while the image is on view will cause the EGA/VGA version of color to enter cursor inquiry mode if a Microsoft- compatible mouse is present. Be sure that the memory- resident driver for the mouse has been loaded prior to running COLOR, otherwise it will ignore all attempts to enter inquiry mode. When inquiry mode has been entered, a cross-shaped cursor will pop up onto the center of the screen and a column and row indicator will appear in the lower left-hand corner. These indicators will always show the correct image position, even when the image is expanded. Note that they will also change color if the cursor is moved off the image. If the image is an unpacked binary image, you may press the left mouse button to inquire about the value (z) at that point. If it is not an unpacked binary image, COLOR will ignore the left mouse button. When an inquiry is made, the mouse will freeze at that position and display the correct value. To unfreeze the display, press the left mouse button a second time and move the mouse. To exit inquiry mode, press the right mouse button. 2. Pressing the letter "d" while the image is on screen will cause the EGA/VGA version of color to enter on-screen digitizing mode. It will first ask for the type of feature to be digitized. Press "P" for point, "L" for line, or "A" for area (polygon). This question will only be asked the first time a request to digitize has been made. COLOR will next ask for an identifier for the feature to be digitized. Enter an integer number followed by the return key. (Only integer identifiers may be used with on-screen digitizing - use a values file to assign real number values to any identifier). If the feature type is point, you will actually be asked for the start identifier, after which it will automatically increment the identifier with each point digitized. In this manner, many points may be digitized at once. For lines or polygons, only a single feature is digitized at one time. You may now begin digitizing. Move the cursor to each point in the feature and press the left mouse button to cause it to be digitized. The feature will be drawn on the screen as you digitize. To finish digitizing the feature (or the set of features in the case of points), press the right mouse button. You will then be asked whether you wish to keep that feature. If there have been any errors, indicate "no" by pressing "n". Note that in the case of polygons, you do not need to close a polygon yourself. Pressing the right mouse button to stop digitizing will automatically close the polygon by joining the first and last points. After you have digitized a feature, you will be returned to normal image viewing mode. To digitize another feature, simply press "d" again. You may do this as many times as you wish. Once you finally exit COLOR (by pressing return while in normal image viewing mode) you will be asked to give a name and a title to the vector file you have created. If you decide you do not wish to keep your digitized data, you can respond by giving the name "quit". 3. Pressing a "v" while the image is on screen will cause the EGA/VGA version of COLOR to overlay a vector file onto the image. It is assumed that the image and vector files are coincident (ie., that the minimum and maximum X and Y values listed for the vector file coincide with the minimum and maximum columns and rows of the image on display). COLOR will first ask for the name of the vector file to be displayed. It will then ask for the color code to be used. Specifying an integer number between 0 and 15 will cause all features in the vector file to be given an identical color using the color index chosen. If, however, you simply press return (or enter any text string such as "none"), COLOR will then use the identifiers of the vector features as the means of assigning colors (ie., an identifier of 6 will cause the feature to be given color 6). Identifiers greater than 15 will be given color 15. Negative identifiers will be assumed to be "negative" space associated with the previous feature (eg., holes within polygons). Any number of vector files may be overlaid onto an image. 4. Palette files contain instructions on the formation of specific color schemes. Palette files must have a ".pal" extension and should be located in your IDRISI directory (rather than your data directory). For EGA graphics cards, the palette file should be a simple ASCII file with two columns separated by one or more spaces. The first column should contain the numbers 0 through 15. These are the legend category numbers. The second column should contain EGA color index numbers which range from 0-63. If a VGA card is in use, the palette file should be a simple ASCII file with four columns of numbers separated by one or more spaces. Again, the first column should contain legend category numbers from 0- 15. The second column specifies the strength of red in the color being specified, using an integer number from 0 (none) to 63 (full strength). The third and fourth columns specify the strength of green and blue respectively, again using this 0-63 numbering system. As an illustration of the use of palette files, here are two examples of palette files, one for the EGA and one for the VGA, that will each produce the same color sequence as the IDRISI default color set. EGA example: VGA example : 0 0 0 0 0 0 1 8 1 0 0 23 2 1 2 0 0 47 3 9 3 0 0 63 4 41 4 23 0 63 5 13 5 47 0 63 6 45 6 63 0 63 7 37 7 63 0 47 8 44 8 63 0 23 9 36 9 63 0 0 10 52 10 63 23 0 11 38 11 63 47 0 12 54 12 63 63 0 13 22 13 47 63 0 14 50 14 23 63 0 15 2 15 0 47 0 If you create a new palette that you use very often, you can give it a special name to allow it to be chosen by default. When COLOR asks for the name of the palette file, simply pressing return will cause it to look for a file named "userdef.pal". Therefore simply give your favorite palette file this name. 5. The EGA/VGA version of color can display only 16 colors at once. In the case of EGA adaptors, these 16 can be chosen from a total of 64 possible colors. With VGA adaptors, however, they may be chosen from 262144 possibilities. As a result, very smooth color transitions are possible with the VGA, yielding images of almost photographic quality. The VGA adaptor also has a resolution of 640 x 480 while the EGA supports 640 x 350. 6. The EGA color index numbering scheme is quite complex. Assuming that the color intensities can be designated by the numbers 0 (least) to 3 (most), and that each color is composed of three primaries in varying intensities in the order Red Green Blue, then Color numbers 0 through 7 have the following characteristics: Color 0 : 0 0 0 Black 1 : 0 0 2 Medium Blue 2 : 0 2 0 Medium Green 3 : 0 2 2 Medium Cyan (Blue-Green) 4 : 2 0 0 Medium Red 5 : 2 0 2 Medium Magenta (Red-Blue) 6 : 2 2 0 Medium Yellow 7 : 2 2 2 Medium Grey To this basic set, the lower intensities of Colors 1 through 7 are added to yield the remaining colors. For example, Color 8 results from adding Dark Blue (0 0 1) to Black (0 0 0) to yield Dark Blue. Color 9 results from adding Dark Blue to Medium Blue (0 0 2) to get Light Blue (0 0 3), and so on, as follows : Colors 8-15 Add Dark Blue (0 0 1) 16-24 Add Dark Green (0 1 0) 24-31 Add Dark Cyan (0 1 1) 32-39 Add Dark Red (1 0 0) 40-47 Add Dark Magenta (1 0 1) 48-55 Add Dark Yellow (1 1 0) 56-63 Add Dark Grey (1 1 1) Thus Color 63 which adds Dark Grey (1 1 1) to the Medium Grey of Color 7 (2 2 2) yields White (3 3 3). Confused? Reasonably enough! Thankfully, the VGA adaptor avoids all of this through the use of standard RGB notation. 7. In the display of legend text, only the first 12 characters will be displayed. CGA VERSION : COLORCGA.EXE Version 3.04 Operation : This version of COLOR works with all CGA compatible adaptors, including most LCD laptop displays. It has two resolution modes -- 160 x 100 resolution with 16 colors and 320 x 200 resolution with 4 colors. Since the CGA only supports 4 colors simultaneously, the 16 color mode is produced by means of dither patterns. The module should be renamed COLOR.EXE for use (this is done automatically during installation). COLOR begins by asking for the name of the image to be displayed, followed by the resolution to be used.It then asks for the palette choice. These are the standard IBM CGA palettes numbered 0-3. It will then ask if you wish a legend and the expansion factor to be used. For the expansion, enter an integer number (eg., 2, 3, 4 etc). If the image is too large to fit into the available display space, you will then be prompted to select the start and end columns and rows for the displayed image. After the image is displayed, the normal text screen can be restored by pressing return. Notes : 1. In the display of legend text, only the first 6 characters will be displayed. 2. None of the interactive query, digitizing and vector overlay features of the EGA/VGA version are supported by the CGA version. HERCULES MONOCHROME : COLORHGA.EXE Version 3.05 Operation : This version of COLOR works with all Hercules monochrome compatible adaptors. Because this is a monochrome device with no grey- level support, dither patterns are used as a substitute for color. It has two resolution modes -- a 240 x 174 resolution mode in 7 shades of grey and a 180 x 116 resolution mode in 13 shades of grey. The module should be renamed COLOR.EXE for use (this is done automatically during installation). COLOR begins by asking for the name of the image to be displayed, followed by the resolution to be used.It will then ask if you wish a legend and the expansion factor to be used. For the expansion, enter an integer number (eg., 2, 3, 4 etc). If the image is too large to fit into the available display space, you will then be prompted to select the start and end columns and rows for the displayed image. After the image is displayed, the normal text screen can be restored by pressing return. Notes : 1. In the display of legend text, only the first 12 characters will be displayed. 2. None of the interactive query, digitizing and vector overlay features of the EGA/VGA version are supported by the Hercules version. General Notes : 1. If your image appears as primarily a single color, you may have forgotten to use the "a" command line parameter. 2. When autoscaling is off (the default), real number images cannot be viewed. 3. When autoscaling is off (the default), integer codes less than zero are forced to be zero, and integer codes greater than the largest color value are forced to the maximum. Error Messages : 1. Without autoscaling, input file must contain integer color codes. - The image you are trying to view contains real numbers. Use the autoscaling feature to view this image by specifying the "a" command line parameter (see above). Otherwise, use STRETCH to create a properly scaled set of color index numbers. 2. Palette file not found. - palette files should be stored in your IDRISI program directory and should be named with a ".pal" file extension. 3. This module requires either an EGA or VGA graphics system. - Generated by the EGA/VGA version if an EGA or VGA board is not present 4. Your EGA card contains only 64K of memory (128K minimum required). 5. Monochrome EGA not supported DESCRIBE Version : 3.01 Purpose : DESCRIBE allows one to display the contents of an IDRISI documentation file on the screen. It achieves the same end as using the TYPE operating system command. However, it has the advantage that it reads the IDRISI environment file to determine the drive, directory and file extension to be used, thus greatly simplifying this operation. Command Line Parameters : The name of the documentation file to be described may optionally be indicated as a command line parameter. See below. Operation : DESCRIBE asks simply for the name of the data file to be described. It will then look for all instances of image, vector and attribute values files that use that name, and describe each in turn. DESCRIBE can also be run by specifying the file name as a command line parameter. Simply type the word DESCRIBE followed by the name of the file to be described (without any drive, directory or file extension information). For example, if one wished to describe the contents of an image named SOILS (stored as SOILS.IMG with an accompanying documentation file of SOILS.DOC), one would simply type the following : DESCRIBE SOILS The contents of the documentation file for soils will then appear on the screen (with pauses if extensive legend material is present). Notes : 1. If you are working on a floppy-disk based system, you should be aware that the installation procedure you undertook in Section 3 placed a copy of DESCRIBE onto all floppy disks. Error Messages : 1. Invalid character (".") in ... - File names should be entered without an extension and should not contain the "." character. DOCUMENT Version : 3.00 Purpose : DOCUMENT is a Data Management Module designed to create and update the documentation files that accompany IDRISI image, vector and attribute values files. Command Line Parameters : Normally, DOCUMENT is used to document image files. However, a vector file can be documented by specifying a "v" after the DOCUMENT command (ie., "DOCUMENT v") while an attribute values file can be documented by entering an "a" command line parameter (ie., "DOCUMENT a"). Operation : DOCUMENT first asks for the name of the image, vector or values file to be documented (remember to use the "v" command line parameter if it is a vector file or an "a" command line parameter is it is a values file). If a documentation file for the image, vector or values file does not exist, DOCUMENT will immediately ask for the information it needs as follows : A. Image Files 1. A title. 2. The data type (integer, real or byte). 3. The file type (ascii, binary or packed). 4. The number of rows. 5. The number of columns. 6. The ground width of each cell (in the x direction). 7. The ground height of each cell (ie., the y direction length rather than vertical height). 8. The number of legend categories. Enter a zero for no legend. If a number greater than zero is identified, you will be prompted to enter the legend captions one at a time. DOCUMENT then automatically calculates the minimum and maximum values in the image file. B. Vector Files 1. A title. 2. The data type of the identifiers (integer, real). 3. The file type (ascii, binary, ascii column or binary column). The columns formats consist of a single column of numbers in the same manner as the older Version 2 format (see GETTING STARTED). The non-column formats are the usual Version 3 formats as detailed in the GETTING STARTED section. 4. The object type (point, line, polygon or chain). The chain type is not used in this release, but will be important when a full vector ring is added. 5. The coordinate span (the ground distance represented by a coordinate difference of one). See GETTING STARTED and the description sheet for DIGITIZE for further information. 6. The units of this measurement (m,ft,mi,km,deg or rad). DOCUMENT will then calculate the minimum and maximum X and Y coordinates that exist within the file. C. Attribute Values Files 1. A title. 2. A subtitle. 3. A data description. This is usually used to indicate the measurement units of the attribute data. 4. The classification type. This can be left blank (as can any of the entries so far) if necessary since it is intended more as an annotation for values files prepared for choropleth mapping (a capability that will be incorporated in a future release of IDRISI). 5. The number of classes. Enter a zero if the data are raw data (not classified). 6. The data format (integer, real or string). The string format is used when the file contains names to describe the feature identifier codes. 7. The number of legend categories. The legend serves exactly the same purpose as that attached to image files. Enter a zero if no legend is to be stored. If a number greater than zero is identified, you will be prompted to enter the legend captions one at a time. DOCUMENT will then automatically calculate the minimum and maximum data values which occur within the file. In all cases, DOCUMENT will next present a display of the current documentation file parameters. If a documentation file had already existed, this would be the first display you would see. Any parameter can be changed by typing the capitalized letter of that parameter, after which a new value will be requested. If the legend is specified for updating, the user will be requested for the category number to be updated (or a zero [or return] to quit). New categories can be added by specifying a number greater than the currently existing number of categories. Legend captions can continue to be added until a zero is entered, at which the summary display will appear again. Legend captions can be deleted by entering blank lines. In this fashion, a legend can be completely erased, one caption at a time. Experiment --it's easier than this description would make it seem! When the summary display is satisfactory, simply press return to save the updated (or new) documentation file. Notes : None Error Messages : 1. Non-integer number of points associated with a feature. - This message may appear during the documentation of a vector file. It occurs when it encounters a vector feature (point, line, polygon) that states a non-integer value for the number of points (x,y pairs) that define it. The most likely cause of this is a corrupted vector file in which the number of points for a feature (the second line in a feature record line group) has been omitted, thus causing DOCUMENT to read the first coordinate value (a real number) for this value. Check your vector file, and make sure that each feature has an identifier and number-of-points line, followed by the correct number of x,y pairs. 2. Units unrecognizable ... please enter either m,ft,mi,km,deg or rad. ENVIRON Version : 3.00 Purpose : A System Operation Module that is used to change the default IDRISI operating environment. Command Line Parameters : None Operation : ENVIRON starts by presenting a summary of the current operating environment, with each parameter numbered. Parameters are changed by selecting the appropriate number, after which the new value is requested. To exit, simply press the return (or enter) key. Notes : 1. The default data disk drive is the drive that IDRISI modules use when reading or writing data. 2. The default data path should begin and end with a backslash. A case of "no path" can be entered by simply pressing return when asked to specify the new path name. An entry of "none" will then appear in the summary. 3. File extensions are the three letter codes (eg. ".img") which are attached to file names to help identify their contents. By default these are set to ".img" for image files, ".doc" for image documentation files, ".vec" for vector files, ".dvc" for vector documentation files, ".val" for attribute values files, and ".dvl" for values documentation files. They may be changed, however, to any convention of preference. 4. Cell units are the linear units being used to indicate the width (x) and length (y) of cells. Acceptable units are "m" (meters), "ft" (feet), "mi" (miles), "km" (kilometers), "deg" (degrees) or "rad" (radians). 5. The banner, dialogue and message colors can all be changed by specifying the standard IBM color numbers which range from 0 - 15 (see the COLOR description sheet in this section). 6. The digitizer and plotter ports refer to the serial ports (eg., com1, com2, etc) to which these devices (if present) are attached. Error Messages : 1. Response not understood. Enter a single digit choice. - Response could not be interpreted. Enter only a single digit without any other characters (eg. no decimals). 2. Enter drive as a single letter. - Disk drive incorrectly entered. Enter as a single letter. 3. Invalid path descriptor. - path descriptor must begin and end with a backslash ("\"). 4. Invalid file extension. - the extension must begin with a period and consist of no more than three characters (eg : ".doc"). 5. File extension too long. - no more than three characters may follow the period. 6. Enter a number between 1 and 14, or press <return> to quit. - the choice you made is not on the menu. HISTO Version : 3.05 Purpose : HISTO produces frequency histograms of the cell values in IDRISI images and query files. HISTO creates histograms by dividing the data range (maximum - minimum) into classes with a width as specified by the user. The frequency within each class is then tabulated. Both graphic and numeric output options are available. Command Line Parameters : Placing a "p" after the HISTO command (eg., "HISTO p") will cause output to be directed to the printer rather than the screen (the default). Non-graphic characters will be used (see the Operation Section on how to print the true graphic version). This version also understands the older "s" command line parameter to direct output to the screen (this is the default now anyway). Operation : HISTO first asks for the name of the image to be analyzed. It then presents the minimum and maximum values in the image and asks if they should be reset (for the purposes of creating the histogram only -- the image documentation file will not be changed). The reason why one would change these is to set where the first histogram class will begin and the last will end. Whenever the minimum can be reasonably reset to zero it is advantageous to do so, since a simple relationship will then exist between histogram classes and data values. If the entire range of values is desired, the maximum specified should exceed the actual maximum since the histogram maximum excludes all values greater than or equal to that value. If the user responds "y" (yes), HISTO asks for the new minimum and maximum values, after which the class width is requested. Finally, HISTO asks whether graphic or numeric output is desired. If graphic output is chosen, the result will depend on whether an EGA or VGA graphics adaptor is present. If it is, a true graphics plot will result. If not, a character- based graphic plot will result with histogram bars being identified by class number only. If you are uncertain about the data ranges of these classes, run HISTO a second time, and ask for numeric output -- numeric output specifies the data ranges for each class. In addition to the frequency of cells within each class, numeric output also lists the proportion of all cells falling within a class, the cumulative frequency, and the cumulative proportion (cumulative relative frequency). If you have an EGA of VGA adaptor, and one of the supported graphics printers (see Section 2), you can also produce a plot of the graphic histogram by pressing the letter "P" while viewing the graphic. Both output types list the class width, histogram maximum and minimum, actual maximum and minimum and mean. Graphic output also lists the Y-ax is scaling factor since it may be necessary to indicate more than one cell by each histogram bar symbol. For example, a Y- axis scaling factor of 10 indicates that each graphic symbol represents 10 cells. Also on the graphic output, classes below the actual minimum and classes above the actual maximum are not shown. Empty classes therefore can be assumed to indicate cases where the frequency is greater than zero, but less than the Y-axis scaling factor. For example, if the Y-axis scaling factor is 10, an empty class indicates a frequency between 1 and 10. Notes : 1. If you are confused by this explanation, experiment. The logic is actually quite simple. While you are learning to use HISTO, you may want to produce numeric outputs for each graphic output you produce. The pair will leave no ambiguity. 2. Histogram maximum and minimum values do not have to include the actual data range, but can focus on a sub-range. The summary at the bottom of the histogram will indicate the number of cells not included. The mean will only be calculated from the values within the specified histogram range (it will indicate the number used). 3. HISTO can accommodate a maximum of 1000 classes (producing a histogram more than 15 pages long -- more than adequate for most applications). Error Messages : None ORTHO Version : 3.xx (Different versions exist for various display adaptors. During Installation, the version chosen (eg. ORTHOEVA.EXE) is normally renamed to be ORTHO.EXE. The filename of each version, as it is distributed, is listed with each of the specific device descriptions below.) Purpose : ORTHO displays and prints "three-dimensional" orthographic perspective displays of IDRISI images. Normally, the kinds of images that would be displayed are surfaces, although any IDRISI image may be viewed with this routine. Additionally, with the EGA/VGA version of ORTHO, a second color image may be "draped" onto the top of the surface displayed. EGA/VGA VERSION : ORTRHOEVA.EXE Version 3.03 Command Line Parameters : With the EGA/VGA version of ORTHO, a "p" command line parameter can be used to direct output to a printer (ie., "ORTHO p"). In doing so, ORTHO will output to the screen while it is composing the printed copy in memory. Only after the plot is finished will output to the printer begin. Operation : ORTHO first asks for the name of the image to be displayed. It will then ask if you wish to drape a second image on top of that surface. If you answer yes, it will ask for the name of the image to be draped and whether you wish to plot a legend. The draped image must be pre-configured into a directly displayable format -- ie., it must contain integer (or byte) values within a range of 0-15. Use STRETCH or RECLASS to create such an image. Like COLOR, ORTHO will display a full text legend and will create a title and subtitle based on the title of the two images involved. In order to do so, ORTHO will also ask for the color palette to be used in displaying the draped image. In addition to a set of standard options (as in COLOR), a user-defined palette may be used. The procedure here is exactly the same as described for COLOR. ORTHO will next display the minimum and maximum values in the image and ask if you wish to alter them. Indicating "n" (for no) or simply pressing return will cause the image minimum to be used as if it were the base level (ie., sea level). Altering those values will allow you to change the level of the base (by lowering the minimum) and the amount of exaggeration (by raising the maximum). However, the exaggeration factor can be controlled directly as outlined below. ORTHO next asks for the view direction, the viewing angle, and the vertical exaggeration factor. Pressing return in each case will choose the defaults of 45 degrees, 30 degrees and 1.0 respectively. The view direction is the azimuth of your look direction (ie.,the direction to which your eye is pointing). The default is 45 degrees (looking north-east), and will thus cause you to be looking at the southwest corner of the image as your closest point. The view direction may be changed continuously by specifying a decimal degree value between 0 (north) and 90 (East). Other directions may also be achieved by using TRANSPOS before viewing to rotate the entire block by 90 degree increments. For example, to view the image with a look direction of 315 degrees (toward the northwest), use TRANSPOS to rotate the block 90 degrees counterclockwise and then use the default viewing direction in ORTHO. The viewing angle refers to the inclination of the eye to the image plane. The default is 30 degrees above the image plane. However, it may be changed continuously (in decimal degrees) between 0 (a horizontal view that would allow one to look at the image in profile) and 90 degrees (a vertical view that would look quite boring -- the image would look like a flat venetian blind). Finally, the vertical exaggeration factor can be used to alter the degree of exaggeration in the display. ORTHO is intended for non-metric visualization only, and thus employs no metric scaling. The degree of exaggeration in a view will depend upon several factors. Using the image minimum as the base value will lead to the greatest amount of exaggeration as will a viewing angle of 0 (horizontal profile). Changing the minimum to lower values and higher viewing angles will both lead to reduced amounts of exaggeration. Perhaps the easiest way of regulating the degree of exaggeration, however, is to set the base level and viewing angle to meet logical and perspective requirements and then (after looking at the display), to use the exaggeration factor to alter a subsequent plot. The exaggeration factor is a multiplicative adjustment. Thus specifying a value of 0.5 will halve the amount of exaggeration while a value of 2 will double it. Any positive value may be specified. Experiment! Printing : The EGA/VGA version of ORTHO can be redirected to a printer using the "p" command line parameter (see above). Note that all colors are treated as black except color 0. As a result, printing of draped images is not supported. A screen capture and printing utility such as PIZAZZ Plus (tm) or DELUX PAINT II (tm) is recommended for printing draped ORTHO images. CGA VERSION : ORTHOCGA.EXE Version 3.01 Command Line Parameters : none Operation : ORTHO first asks for the name of the image to be displayed after which it will indicate the minimum and maximum values in the image and ask if you wish to alter them. Indicating "n" (for no) or simply pressing return will cause the image minimum to be used as if it were the base level (ie., sea level). Altering those values will allow you to change the level of the base (by lowering the minimum) and the amount of exaggeration (by raising the maximum). However, the exaggeration factor can be controlled directly as outlined below. ORTHO next asks for the view direction, the viewing angle, and the vertical exaggeration factor. Pressing return in each case will choose the defaults of 45 degrees, 30 degrees and 1.0 respectively. The view direction is the azimuth of your look direction (ie.,the direction to which your eye is pointing). The default is 45 degrees (looking north-east), and will thus cause you to be looking at the southwest corner of the image as your closest point. The view direction may be changed continuously by specifying a decimal degree value between 0 (north) and 90 (East). Other directions may also be achieved by using TRANSPOS before viewing to rotate the entire block by 90 degree increments. For example, to view the image with a look direction of 315 degrees (toward the northwest), use TRANSPOS to rotate the block 90 degrees counterclockwise and then use the default viewing direction in ORTHO. The viewing angle refers to the inclination of the eye to the image plane. The default is 30 degrees above the image plane. However, it may be changed continuously (in decimal degrees) between 0 (a horizontal view that would allow one to look at the image in profile) and 90 degrees (a vertical view that would look quite boring -- the image would look like a flat venetian blind). Finally, the vertical exaggeration factor can be used to alter the degree of exaggeration in the display. ORTHO is intended for non-metric visualization only, and thus employs no metric scaling. The degree of exaggeration in a view will depend upon several factors. Using the image minimum as the base value will lead to the greatest amount of exaggeration as will a viewing angle of 0 (horizontal profile). Changing the minimum to lower values and higher viewing angles will both lead to reduced amounts of exaggeration. Perhaps the easiest way of regulating the degree of exaggeration, however, is to set the base level and viewing angle to meet logical and perspective requirements and then (after looking at the display), to use the exaggeration factor to alter a subsequent plot. The exaggeration factor is a multiplicative adjustment. Thus specifying a value of 0.5 will halve the amount of exaggeration while a value of 2 will double it. Any positive value may be specified. Experiment! Printing : The CGA version of ORTHO also supports printed output to Epson/IBM printers. Pressing any number from 0 to 6 while the perspective plot is being used will cause the currently displayed image to be printed using that printer graphic mode. Do not press a return after pressing the number (since a return will cause the image to disappear). Graphics mode 0 usually leads to a full-page image while mode 1 leads to a reduced image. Experiment! A screen capture and printing utility such as PIZAZZ Plus (tm) or DELUX PAINT II (tm) is also recommended for printing ORTHO displays. Notes : 1. Titles may also be added to CGA ORTHO displays. To add a title line, press a "t" (but do not terminate it with a return) immediately followed by the title characters. The title itself should be terminated with a return. This process can be repeated to produce multiple-line titles. Note that titles added in this way will not be printed unless a screen capture utility is used. HERCULES VERSION : ORTHOHGA.EXE Version 3.01 Command Line Parameters : none Operation : The operation of the HERCULES graphics adaptor version of ORTHO is identical to that required for the CGA version (including printing). See above. OVERLAY Version : 3.00 Purpose : An analytical module that produces a new image from the data on two input images. New values can result from any of the following operations on the two input images (referred to as the "first" and "second" images during program operation) : [1] Add : Corresponding pixels from the two images are added. [2] Subtract : Pixels from the second image are subtracted from those of the first. [3] Multiply : Corresponding pixels from the two images are multiplied. [4] Ratio : Pixels from the first image are divided by corresponding pixels from the second. [5] Normalized Ratio : Divides the subtraction of the two images by their addition. Commonly used in the derivation of vegetation indices from remotely sensed data. [6] Exponentiate : Raises pixels in the first image to the power of corresponding pixels in the second. [7] Cover : Covers pixels of the second image with those of the first, except where the first image has values of zero -- in these cells, the value of the second image shows through. [8] Minimize : Output pixels represent the minimum of those in corresponding positions on the first and second images. [9] Maximize : Output pixels represent the maximum of those in corresponding positions on the first and second images. Command Line Parameters : None Operation : OVERLAY begins by presenting the menu of overlay types discussed above. Select the overlay desired by entering a single digit number. Next, OVERLAY asks for the names of the "first" and "second" images. Pay careful attention to the relationship between these two images in the overlay as indicated in the menu. Finally, OVERLAY asks for the name of the output image (enter a new name up to 8 characters in length), and a title. Notes : 1. Boolean analyses can be undertaken with OVERLAY using binary images. See the GETTING STARTED section on "Analyzing Images". 2. OVERLAY uses standard mixed arithmetic rules in determining the data type of the result. 3. OVERLAY will create an output file in packed binary format only if both input files are stored in this manner. 4. OVERLAY will create a byte output file only if the input files are in byte format and the lowest possible minimum and the highest possible maximum fall with the range of 0-255. If it is determined at the end that an output file predicted to possibly have non-byte values does completely conform to byte limits, a message will be displayed to indicate that compaction to byte format is possible. Error Messages : 1. File types don't match. Output file type is binary. - One input file is in ASCII format while the other is in binary format. This is a warning only, and will not stop operation -- the output file format is set to binary. 2. Note : Output file is suitable for compaction to byte format. - This is an informational note only. After completion, OVERLAY has determined that conversion to byte format (a more compact data format) is possible. Use CONVERT to do so if you wish. PARE Version : 3.00 Purpose : PARE "pares" (ie., removes) the header or trailer from a data file. It is commonly used to convert a foreign data file to an IDRISI image format. Many GIS and image processing systems use a format very similar to IDRISI but attach the documentation to the front of the file (known as a "header"). PARE can be used to remove the header and thus yield a file that IDRISI can read. Command Line Parameters : Using an "n" command line parameter (ie., "PARE n") will cause PARE to expect that the file contains 4-bit integers. A 4-bit integer has a range from 0-15 and requires only half a byte for storage. A half-byte unit is sometimes called a "nybble" - hence the "n" parameter. Operation : PARE begins by asking for the name of the file to be pared, complete with its extension. It will then ask for a new name for the IDRISI image to be produced. You will then be asked for the size of the header to be removed (always specify this in bytes) followed by the number of rows and columns in the image to result. Finally, it will ask for the width and height of each cell and a title for the image to be produced. Notes : none Error Messages : none RECLASS Version : 3.00 Purpose : RECLASS provides facilities for classifying or reclassifying the data stored in images or attribute values files into new integer categories. Classification (or reclassification) is either by equal intervals division of the data range, or by the application of user-defined limits. Command Line Parameters : Specifying an "a" command line parameter (ie., "RECLASS a") will cause RECLASS to expect to reclassify an attribute values file rather than the usual image file. Operation : RECLASS first asks for the name of the file to be classified (or reclassified), followed by a new name (up to 8 characters in length) for the classified result. RECLASS then asks that you choose between : [1] : Equal Intervals Classification [2] : User-defined Classification If option [1] is chosen, you are then presented with the current minimum and maximum values in the input image, and asked whether you would like to change them (only for the duration of this module). It is not uncommon that the minimum and maximum are changed in order that the classes derived start and end at some even value. Thus, for example, for a data set with a minimum of 2 and a maximum of 57, you might change these to be 0 and 60. Answer "y" or "n" according to your choice. If your answer is yes, RECLASS will then ask for the new values to assign to the minimum and maximum. In either case, you will next be asked whether you wish to set the classification by specifying the number of classes to be used or the width of each class. Depending upon your choice you will then be asked either for the number of classes into which the data should be divided or the width to be used in devising each class. Finally, you will be asked for a title. If option [2] is chosen, you will be presented with some brief instructions on entering the limits and values of each class. The sequence is quite straightforward. For each class, you must enter the new value (an integer) of the class. You will then be asked for the range of old values (which may be real numbers) that should be assigned to this new class. The choice of the old values is important. The class will include the lower of the old values but not the higher one (ie., it will include all values equal to or greater than the lower one, and less than the higher one). Values outside the actual data range are permitted, and are a good way of ensuring that all old data values are in fact reclassified. For example, suppose the old data values range from 11 to 51, and it was intended to create two new classes: the first containing all old values less than 30 and the second containing all old values greater than 30. Choosing ranges from 11 to 30 for the first and 30 to 51 for the second would be incorrect since the value 51 would not have been included in the higher class. Ranges from 11 to 30 for the first, and 30 to 52 would work. Continue entering ranges like this until all classes have been defined (the order of defining classes is not important). When you wish to finish, enter a "-1" as the new value of the next class. You will then be asked for the title of the new image. Notes : 1. There is no real difference between classification and reclassification with respect to the operation of this module. Error Messages : 1. Unclassified value greater than 32767 set to 32767. - This is a warning only that may occur with input images containing real numbers. If a cell is not reclassified (because of the reclassification ranges specified), it normally is left unchanged. However, integer numbers may not exceed 32767. Therefore, an unclassified (ie., unchanged) value which exceeds this must be forced to conform to the limits of the integer data type (all output images from RECLASS are integer). 2. Unclassified value less than -32767 set to -32767. - This is a warning only that may occur with input images containing real numbers. If a cell is not reclassified (because of the reclassification ranges specified), it normally is left unchanged. However, integer numbers may not be less -32767. Therefore, an unclassified (ie., unchanged) value which is less than this must be forced to conform to the limits of the integer data type (all output images from RECLASS are integer). 3. Specified width does not fit evenly within the range. Max adjusted. - This message occurs when equal intervals classification is achieved by specifying the width of the classes, and the range from minimum to maximum does not contain an integer number of these widths. In this case, the maximum is adjusted upwards automatically. This is a warning only, and does not halt module operation. 4. Attribute values file contains string data. - Only attribute values files which contain numeric data can be submitted to this module. STRETCH Version : 3.00 Purpose : STRETCH undertakes a linear rescaling of image values to fall with a range from 0 to a user-specified upper limit in preparation for the IMAGE and COLOR modules. Command Line Parameters : None Operation : STRETCH first asks for the name of the image to be stretched and a new name (up to 8 characters in length) for the resulting image. Next, the upper and lower bounds of the stretch are requested. The upper bound and all values greater than it will be forced to have the value of the upper limit (to be specified momentarily) while the lower bound and all values less than it will become 0. Numeric values should be specified for both. However, the responses "max" and "min" may also be entered to indicate the maximum and minimum values of the image (generally not a good option). Commonly, the upper and lower bounds are set such that 5% of pixels fall outside each of these limits. HISTO can be used to determine these cut-off points (looking for the 0.05 and 0.95 points in the cumulative proportion column of the numeric output is perhaps the most direct approach). Without doing this, contrast may suffer for very skewed data sets. STRETCH then asks for the number of levels required. If the image is being stretched for IMAGE, 32 should be specified (yielding values from 0 - 31), whereas if it is being stretched for COLOR, the number will vary from 4 to 8 (for the CGA version) to 16 (for the CGA,EGA or VGA versions). STRETCH finishes by asking for a title for the output image. Notes : None Error Messages : 1. Output is suitable for conversion to a byte or packed format. - To aid in the integration of user-created modules, IDRISI modules always strive to maintain the data and file types of input images. This is not, however, always the best choice for non-programming application. In this case, because of the nature of the input file, an integer output file has been created. However, the file is suit- able for data compaction using either a byte binary or a packed binary format. Use CONVERT to change to one of these types if you wish. WINDOW Version : 3.00 Purpose : To extract a sub-image from the original as a new image. Command Line Parameters : None Operation : WINDOW first asks for the name of the original image followed by a new name (up to 8 characters) for the resulting window image. WINDOW then asks for the column and row numbers of the upper-left corner of the sub-image window, followed by the column and row numbers of the lower-right corner of the window. Remember that columns and rows are numbered from 0, with the rows progressing from the top to the bottom. WINDOW next confirms the number of rows and columns of the output sub-image (the window image), after which it asks for a title for the new file. Notes : 1. Remember that columns and rows are numbered from 0, with the rows progressing from the top to the bottom. 2. To extract a rectangular region that is not oriented in the same direction as the grid of the original image, use the RESAMPLE module. Error Messages : None Some Notes on Printers with IDRISI: The three letter codes for each of the printer types are as follows (if your printer isn't listed here then refer to the instructions on how to create a new driver for the device you own) : DEC LA34, LA50, LA75, LA100, LN03 : dec Diconix 150 Ink Jet : dcx IBM Graphics Printer or compatible : ibm IBM Proprinter or compatible : pro Epson FX-85 or compatible : eps Epson LX-810 or compatible : lx8 Hewlett Packard PaintJet/Laserjet : hpj Hewlett Packard DeskJet : hdj Toshiba 321/321SL or compatible : tos Creating New Printer Drivers You will only need to read this section if you have already installed IDRISI and are experiencing printer problems. Otherwise, you may skip this section entirely. Most IDRISI modules that produce printer output share a common printer "driver". However, it may be that you have a printer that does not match one of these types. If so, here is what you should do: 1. Try each of the supplied printer drivers in turn to see if this alleviates the problem. You can do this by repeating the steps for installing your printer as outlined in Section 3. It is often the case that a printer may emulate another printer of a different make. For example, many dot matrix printers emulate the IBM Proprinter driver supplied here. If you still have no success, then read on ... 2. The printer drivers supplied with IDRISI are simple text (ascii) files that contain the escape sequences used by printers to control their characteristics. These escape sequences are used by all printing modules other than the IMAGE module (IMAGE is a special graphics module that cannot be accommodated so easily). Therefore, you simply need to create a new driver file that contains the escape sequences which are appropriate for your printer. IDRISI printer drivers consist of 16 lines of text, with the first 14 characters of each line being used for annotation and the remaining 66 characters being used to indicate required information. Here is a printout of the exact contents of the Proprinter driver supplied with IDRISI (named PRCOD- ES.PRO) printer : IBM Proprinter normal : \027F\027W0\018\0273\036\027@ bold : \027E expanded : \027W1 compressed : \015 elite : \027: 6 lpi : \0273\036 12 lpi : \0273\018 normal chrs : 80 elite chrs : 96 conden chrs : 137 resolution : 11 memory used : conventional orientation : portrait Def. width : 7.5 Def. height : 5.625 Notice that the first 14 characters (up to and including the blank after the colon) are for descriptive purposes only -- they are in fact ignored by all printing modules. The first line lists a name for the printer, up to 14 characters in length. You will see this name appear on the top banner of all modules which use the driver (names longer than 14 characters will be truncated). Line 2 indicates the codes required to place the printer into a "normal" state. In some cases, a single command will do this, but the critical issue is that the commands listed should set the characters to an unbolded condition, the print pitch to 10 characters per inch, and the line spacing to single spaced (6 lines per inch). In the example shown here, multiple codes were required to achieve this (you will be instructed on how to specify codes in a moment). Line 3 indicates the codes required to start bold printing and Line 4 indicates the codes required to produce expanded lettering (eg., 6 characters per inch). You may specify the code for any expanded size you wish. Line 5 indicates the codes required for compressed (or condensed) characters -- typically 16.5 - 17 characters per inch. Line 6 indicates the codes required for elite (12 characters per inch) pitch. Lines 7 and 8 indicate the codes for setting the printer to single spacing (6 lines per inch) and half-spacing (12 lines per inch). Lines 9, 10 and 11 indicate the number of characters that can be printed in a complete line of normal (10 cpi), elite (12 cpi), and condensed (eg 17 cpi) characters respectively. Line 12 indicates the resolution (dpi). Line 13 indicates the type of memory available to the printer. Line 14 indicates the orientation of the paper, and lines 15 and 16 describe the default size of the image to be printed. Lines 2 - 8 all require that control codes (such as escape sequences) be indicated after the descriptive title. All printers send control codes in the form of ascii characters (see BASIC CONCEPTS if you are unfamiliar with this term). Some ascii characters are printable (such as the letters on this page); others are not. Therefore IDRISI uses a convention for specifying ascii codes that is identical to that used by many other software packages, such as LOTUS 1-2-3 and REFLEX. To specify an ascii character, you can type either its printable representation, or the ascii code number as a three digit number preceded by a backslash (\). For example the ascii code for the capital letter A can be specified either by typing the letter "A", or by typing "\065" (since ascii code 65 is used to represent the letter A). Many printer codes require the "escape" ascii code which is code number 27. It would be represented in an IDRISI driver as \027 (note again that the code must be specified by three numerals after the backslash). For example, the code to turn on bold printing on an IBM Proprinter is "esc-E" (escape E), which you can see is coded as \027E. It could equally have been coded as \027\069 (since the letter E is ascii code 69). Note that no spaces should be placed between multiple codes and that upper/lower case does matter when specify ascii codes directly. Also, note that all codes for a given attribute must be no more than 66 characters in total (ie., the combined codes fit onto one line along with the 14 character description). Finally, note that the order in which these code are specified cannot be changed. To create your own customized driver, use a text editor or word processor that can create text file (ascii) output to create a file that looks like the example shown here, but with the codes and parameters required by your own printer (or even simpler, make a copy of one of the supplied drivers such as PROCODES.PRO and then edit it to suit your needs). Your printer manual will list the information for your specific printer. Then, when the driver is complete, copy it to all working IDRISI disks and rename it IDRISI.PRN. All printing modules (other than IMAGE) look for a file named IDRISI.PRN and read it to get the required codes for printer operation. Finally, note that you can modify any printer driver in this fashion to take advantage of such features as special fonts, colored ribbons, and the like. Experiment .. it's harmless. Changing the Text Colors The ENVIRON module can be used to change the colors used for the display of text. Users with monochrome monitors may also wish to alter the display colors since some graphics cards change the brightness of text letters according to the color it is trying to emulate. To alter the colors of ei- ther the banner, the main text, or the error and warning messages, select the appropriate item from the list displayed by ENVIRON and enter a new color value according to the following list. 0 Black 8 Dark Grey 1 Blue 9 Bright Blue 2 Green 10 Bright Green 3 Cyan 11 Bright Cyan 4 Red 12 Bright Red 5 Magenta 13 Bright Magenta 6 Brown 14 Yellow 7 Light Grey 15 White You will notice that all three colors are represented in the ENVIRON display. As a result, you can monitor the changes as you specify them. Note that to return the system to its default colors, simply press return when asked for the new color number.