LAS creates a whole set of computer files related to the processing of its various components. The types and maintenance of some of the more common files used are described in this section. The types of files that may be found within LAS include:
In particular, in the course of processing raster images several types of associated data files are created for each image. When creating these files, the LAS modules use the standard file extensions to implicitly associate these files with the image. Accordingly, LAS considers all files in a directory which have the same filename but different extensions to be asscociated. The existence of separate files for the associated information gives increased flexibility, simplifies the task of accessing the associated information, and allows an image and its associated files to be easily treated as a "data set" for archive and file transfer.
The standard file extensions are listed below. Although files may use any extension, users should avoid using these standard extensions for other types of files.
dltf | Display and lookup table file |
ddr | Data descriptor record (DDR) |
gof.pnt | Graphics overlay point file |
gof.line | Graphics overlay line file |
gof.poly | Graphics overlay polygon file |
gof.anot | Graphics overlay annotation file |
grid | Geometric mapping grid |
his | History |
ibis.tab | LAMS tabular file* |
ibis.graph | LAMS graphics file* |
lt | General purpose labeled table file |
img | Image data file |
mtp | Merged tie point file |
rwt | Resampling weight table file |
proj | Projection definition file |
stat | Statistics file |
tpl | Tie point location file |
tps | Tie point selection file |
txt | General purpose ASCII file |
The formats for each of these files and a complete list of the types of files found in LAS are described in detail in the LAS Programmer's Guide.
LAS images will always have the following information associated with them:
An LAS image file may contain one or more bands. For remotely sensed imagery, the multiple bands generally represent data from different parts of the electromagnetic spectrum; for this reason, they are often referred to as spectral bands. However, the bands of an image file may also represent other types of information, such as the same spectral band viewed at different dates, or various types of geographic information (elevation, landcover class, etc.)
LAS applications modules accept both single-band and multi-band images. A multi-band input image may be either a single image containing more than one band or a logical image composed of a set of images each containing one or more bands. In the latter case, the logical image is specified by joining the names of the individual images by the plus sign, "+".
An optional window specification may be used along with each image name to specify spatial and/or spectral windows. The most general form of this specification is
If spatial-window is omitted, all lines and samples of the image are used. If spectral-window is omitted, all bands are used. If the first two specifications are omitted, the entire image is used. If the spatial-window-type is omitted, the spatial window is assumed to be in image (line and sample) coordinates.
For the default (image line and sample) case, the spatial window specification is given as (SL,SS,NL,NS), where the commas may be replaced by spaces, and the values are
SL | start line number | |
SS | start sample number | |
NL | number of lines | |
NS | number of samples |
Note that LAS numbers lines (rows) from top to bottom, beginning with 1, and samples (columns) from left to right, also beginning with 1.
The spectral window specification is given as a colon followed by integers specifying the bands to be used. This specification may be a list of band numbers separated by commas or blank spaces, or two band numbers separated by a dash, "-", to denote consecutive bands. For example, bands one through four may be specified as ( :1,2,3,4) or ( :1 2 3 4) or ( :1-4).
Section 4.2.2.1 describes the optional spatial-window-type specifier.
Examples:
For a LAS module (such as copy, list, or pixcount) which requires one logical image, specified by parameter IN, as input, the image specification could take any of the following forms:
IN=IMAGE1 IN="IMAGE1(1 1 100 100)" IN="IMAGE1(:1 2)" IN="IMAGE1(1 1 100 100:1 2)" IN="IMAGE1+IMAGE2" IN="IMAGE1(1 1 100 100)+IMAGE2(1 1 100 100)" IN="IMAGE1(:1 2)+IMAGE2(:1 2)" IN="IMAGE1(1 1 100 100:1 2)+IMAGE2(1 1 100 100)" IN="IMAGE1(200,200,100,100:2-5)+IMAGE2(1,1,100,100:3,5)"
The first case uses all lines and samples (rows and columns) of all bands of IMAGE1. The other cases specify various spatial and/or spectral subsets of one or two images. In particular, in the last case, IMAGE1 and IMAGE2 are both multi-band images. The window 200,200,100,100 will be applied to bands 2,3,4, and 5 from IMAGE1 and the window 1,1,100,100 to bands 3 and 5 from IMAGE2. The LAS module will process these six bands as if they all belonged to a single image with 100 lines of 100 samples each.
If the command is entered at the LAS command prompt, the complete command line for copying two bands from each of two different input images into a four-band output image might take the form
COPY IN="IMAGE1(:1 2)+IMAGE2(:2 3)" OUT=IMAGE3For modules that allow or expect multiple logical images -- such as math, composit, and diff -- IN will be a multidimensioned parameter which could take any of the following forms:
IN=(IMAGE1,IMAGE2) IN=("IMAGE1(1, 1, 100, 100)","IMAGE2(1, 1, 100, 100)") IN=("IMAGE1(:1 2)","IMAGE2(:1 2)") IN=("IMAGE1(1 1 100 100:1,2)","IMAGE2(1 1 100 100:1,2)") IN=("IMAGE1+IMAGE2","IMAGE3+IMAGE4") IN=("IMAGE1(:1 2)+IMAGE2(:1 2)","IMAGE3(:1 2)+IMAGE4(:1 2)") IN=("IMAGE1(1 1 100 100)+IMAGE2(1 1 100 100)", + "IMAGE3(1 1 100 100)+IMAGE4(1 1 100 100)") IN=("IMAGE1(1 1 100 100:1 2)+IMAGE2(1 1 100 100:1 2)", + "IMAGE3(1 1 100 100:1 2)+IMAGE4(1 1 100 100:1 2)")The comma between logical images may be replaced with a blank. Note that the "+" at the end of the first line for each of the last two examples is the TAE continuation character; in any other position, it would imply joining the images into a single logical image. Also, when a multi-valued parameter is being entered in tutor mode, the parentheses around the set of values may be omitted, and the separate elements may be entered individually as in the following examples:
IN="IMAGE1(:1 2)" "IMAGE2(:1 2)" IN(1)=IMAGE1 IN(2)=IMAGE2 IN(1)="IMAGE1(1 1 100 100:1 2)+IMAGE2(1 1 100 100:1 2)" IN(2)="IMAGE3(1 1 100 100:1 2)+IMAGE4(1 1 100 100:1 2)"As an example of a complete LAS command which treats the input images as two separate logical images, the command line for adding band 1 of image1 to band 2 of image2, and band 2 of image1 to band 3 of image2 to create the two-band image image3, is as follows:
MATH-ADD IN=("IMAGE1(:1 2)", "IMAGE2(:2 3)") OUT="IMAGE3"
Restrictions and notes:
For the most general form of the image window specifier,
MAX_LAT | = | Maximum latitude |
MIN_LON | = | Minimum longitude |
MIN_LAT | = | Minimum latitude |
MAX_LON | = | Maximum longitude |
UL_Y | = | Map projection Y coordinate at upper-left corner |
UL_X | = | Map projection X coordinate at upper-left corner |
LR_Y | = | Map projection Y coordinate at lower-right corner |
LR_X | = | Map projection X coordinate at lower-right corner |
(not specified) | (default) | ||
"LS" | = | Line/Sample | (default) |
"DEG" | = | Decimal degrees | (Method 2) |
"DMS" | = | DDDMMMSSS.SSSS | (Method 2) |
"PRO" | = | Projection | (Method 3) |
Method 1 (default)
If spatial-window-type corresponds to the default method, then the window will be processed using the standard method described in Section 4.2.2.
Method 2
If spatial-window-type corresponds to Method 2, then:
IN="IMAGE1 (20.3 -120.5 18.6 -118.2 :1 2;DEG)" IN="IMAGE1 (020018000 -120030000 018048000 -118012000:1 2;DMS)" IN=("IMAGE1 (20.3 -120.5 18.6 118.2:1 2;DEG), + IMAGE2 (020018000 -120030000 018048000 118012000:1 2;DMS)")
Method 3
If spatial-window-type corresponds to method 3, then:
IN="IMAGE1 (103400 234560 98650 239640:2 3;PRO)" IN="IMAGE1 (207300 125430 187520 145360;PRO"
Image history information is stored in an associated history file. The image history lists the modules used in the creation of the image. For each LAS module executed, the history includes the date and time at which it was run, the name of the user, the name of the module, and the values of the parameters used.
The history of a LAS image file may be displayed on the terminal by running the module dsphistry. Any computer system utility which will display an ASCII file may also be used, but the displayed information for each module will contain some extraneous record-control information preceding the date field; this should be ingnored. A sample history file is shown below.
Fri Sep 1 16:55:23 1989 UNIX quirk GEOM1P IN="nhap.raw.map" INGRID="nhap.grid" OUT="nhap.reg" WINDOPT="OUT" WINDOW=-- RESAMP="CC" INRWT=-- PCCALPHA=-0.5 BACKGRND=0.0 INTLIM=-- ODTYPE="SAME"Tue Sep 5 08:37:23 1989 UNIX quirk copy IN="nhap.reg(2232,2943,2100,2709)" OUT="nhap.reg.sub" ODTYPE="SAME" ROUNDOPT="TRUNC"
Thu Sep 7 00:36:01 1989 UNIX quirk ISOCLASS IN="nhap.reg.sub" OUT="nhap.reg.iso" OUTSTAT="nhap.stats" STATSRC="NONE"INSTAT=--MAXNUMIT=10CLUSDIST=3.2MAXCLSTD=4.5 SSEPVAL=0.0PRINTIT=11VIEWIT=21MINNUM=30MAXCLUST=24CHNTHR=3.2 PRINT="LP" PRTFILE="ISOCLASS_112544.pr"
Every image has an associated data descriptor record (DDR) file. The DDR file contains five basic types of information. Each time a module is executed, each of the fields is updated. The rules governing how each field is updated is outlined in the LAS Programmer's Guide. The DDR can be displayed using the LAS module dspddr, or each of the fields can be edited by using the LAS module editddr. The five types of information in the DDR are:
Most modules only require valid entries for these items; they will execute successfully even if some or all or the other entries are invalid. There are some modules, however, which require other fields to be valid. For example, concat-auto requires the corner coordinates and related projection information to be valid.
These coordinates are useful in the mosaicking of images and/or retrieval of the projection coordinates of any pixel in the image. For a more detailed description of these fields and how they can be used, refer to the Geometric Registration Overview document.
This is calculated by any module that creates an output image unless the TAE global variable $MINMAX is set to "no."
These fields are used to keep track of acquisition date and time of each band of the image and how it was acquired.
This information includes the last date and time an image was modified and the master line and sample of an image. The last access date and time are useful in deciding which images should be archived. The master line and sample keep track of where pixel (1,1) of the image is relative to an original image from which it may have been extracted as a subset.
Typical output from dspddr for a Landsat image is as follows:
IMAGE NAME:landsat_sample NL:512 NS:512 NB:3 DTYPE:BYTE LAST USED: DATE:26Jan89 TIME:1451:03 SYSTEM:gouldutx PROJ. CODE:(1)UTM Valid:VALID ZONE CODE:11 Valid:VALID DATUM CODE:0 Valid:VALID PROJ. PARM: Valid:VALID A: 0.00000000000000E+00 0.00000000000000E+00 0.00000000000000E+00 B: 0.00000000000000E+00 0.00000000000000E+00 0.00000000000000E+00 C: 0.00000000000000E+00 0.00000000000000E+00 0.00000000000000E+00 D: 0.00000000000000E+00 0.00000000000000E+00 0.00000000000000E+00 E: 0.00000000000000E+00 0.00000000000000E+00 0.00000000000000E+00 CORNER CORR: Valid:VALID ULcorner:3.65777500000000E+06 4.99200000000000E+05 URcorner:3.65777500000000E+06 5.11975000000000E+05 LLcorner:3.64500000000000E+06 4.99200000000000E+05 LRcorner:3.64500000000000E+06 5.11975000000000E+05 PROJ. DIST:2.50000000000000E+01 2.50000000000000E+01 Valid:VALID PROJ. UNITS:meters Valid:VALID INCREMENT:0.00000000000000E+00 0.00000000000000E+00 Valid:INVALID MASTER CORR:0 0 IMAGE NAME:landsat_sample BAND NO:1 MINIMUM:0.00000000000000E+00 Valid:VALID MAXIMUM:1.95000000000000E+02 Valid:VALID DATA SOURCE:landsat4 SENSOR TYPE:mss CAPT. DIRECTION:ascending DATE:23apr86 TIME:1305:55 IMAGE NAME:landsat_sample BAND NO:2 MINIMUM:0.00000000000000E+00 Valid:VALID MAXIMUM:2.33000000000000E+02 Valid:VALID DATA SOURCE:landsat4 SENSOR TYPE:mss CAPT. DIRECTION:ascending DATE:23apr86 TIME:1305:55 IMAGE NAME:landsat_sample BAND NO:3 MINIMUM:0.00000000000000E+00 Valid:VALID MAXIMUM:1.64000000000000E+02 Valid:VALID DATA SOURCE:landsat4 SENSOR TYPE:mss CAPT. DIRECTION:ascending DATE:23apr86 TIME:1305:55
In the process of interpretation and image analysis, some LAS modules use or generate statistics files. These files may or may not be associated with a specific LAS image. LAS statistics files are tree-structured files that contain statistical information derived from the associated images. The types of information contained in statistics files are mean and covariances matrices, correlation data, histograms, and polygon vertices.
When training sites are defined by the mensuration modules, the site polygon vertices are based on entire image coordinates even if the image needed to be subsampled to be displayed. For example, if polygons are selected from a 512 by 512 subsection, starting at image coordinates (1001,2001), the coordinates of any site polygon within that subsection will be in the range (1001,2001) to (1512,2512). This is important to users who wish to edit their polygon vertices.
The data in a statistics file is stored in a data tree structure. The tree structure consists of three levels. The three levels are:
Statistics files are generated by the clustering/classification modules such as isoclass or by the training site definition module put-poly. Note that the statistics files generated by the clustering/classification modules may not have any "sites" for Level 2; they may only have class information at Level 1.
Display and lookup tables files (DLTF) are used within LAS to store lookup tables that are utilized by both the LAS display module (xid) and enhancement modules such as map. A DLTF is a file that defines a sample intensity mapping to be applied to an image. The sample value in an input image is the index to the DLTF, and the contents of the index is the sample value of the output image that would result if the DLTF was applied to the input image. A DLTF can be used with either black and white or color images.
Several types of files are used in performing geometric transformation of image data in LAS. Not all geometric corrections will use all of the file types listed as described in the following list, although some of them will. Additional information may be referenced in the Geometric Registration Overview. The types of files used in geometric registration are:
The "Classic" (non-Xwindows) subset of the Transportable Applications Environment (TAE) also creates and updates a set of files on the computer system. These include parameter files, session histories, PDFs, script files, and the results of batch processing. These files are described in detail in the TAE Command Language (TCL) User's Manual, but several are included in the following list. These files will be created within the user's current working directory during an LAS/TAE session unless the user has explicitly specified another directory (e.g., in a save command during a tutor session).