idrisi taiga gis image processing technical specifications

8/4/2019 IDRISI Taiga GIS Image Processing Technical Specifications

1/12

GIS AnalysisAt the very heart o GIS is the ability to perorm analyses based on

geographic location. Indeed, no other type o soware can provide this.

IDRISI oers a wealth o analytical tools or geographic analysis.

databasequery

Database Query is the most undamental o GIS operations.

RECLASS perorms image reclassication.

OVERLAY perorms nine dierent operations between two images

including add, subtract, multiply, divide, normalized ratio, exponentiate,

minimize, maximize, and cover.

CROSSAB perorms a crosstabulation or a crosscorrelation between

two qualitative maps.

EXRAC calculates summary statistics or a set o input maps.

BREAKOU creates Boolean maps or all categories in an image.

HISO provides a graphic or numeric requency histogram and

statistics o the cell values within an image.

AREA calculates the area in a variety o units o each class in an image.

PERIM calculates the perimeter o each class in an image.

PROFILE creates proles over space or over time.

QUERY extracts pixels designated by an independent mask into asequential le or subsequent statistical analysis.

PCLASS perorms a probability reclassication when the level o

uncertainty in an image is known.

Database Workshop is a relational database manager and provides

the ability to create, edit and analyze database les in IDRISI using

the Microso ADO and Access Jet Engines. Both the Calculate and

Filter operations are supported through the use o Structured Query

Language (SQL).

Image Calculator is an interactive mathematical modeling tool

that allows you to enter a model as a ull algebraic equation using a

calculator-like interace. It also supports mathematical expressions andlogical queries.

mathematicaloperators

IDRISI provides a set o mathematical tools necessary or complete map

algebra.

OVERLAY can perorm nine dierent operations between two images

including add, subtract, multiply, divide, normalized ratio, exponentiate,

minimize, maximize, and cover.

Image Calculator can evaluate models such as this vulnerability mapping analysis using

uzzy criteria in West Arica. The smaller image o market accessibility was produced wit

an anisotropic cost unction that calculates eective distance rom markets recognizing

the rictional eects o landcover, road type, slope and aspect (i.e., whether the slope i

uphill or downhill).

Te IDRISI aiga GIS and Image Processing soware includes nearly 300 modules or the analysis and display

o digital inormation. Te specications or these modules are presented in the same logic as the IDRISI menu

structure within the soware.

IDRISI TaigaGIS and Image Processing Sotware

Technical Specifcations


2/12

SCALAR undertakes arithmetic operations between a constant and a

single image.

RANSFORM can perorm 15 dierent mathematical transorma-

tions on the attributes o a single image including natural logarithms

and antilogs, a logit transormation, reciprocal, square and square

root, absolute value, and all o the trigonometric operations.

Image Calculator is an interactive mathematical modeling tool that

allows you to enter a model as a ull algebraic equation using a calcu-

lator-like interace and supports mathematical expressions and logical

queries.

distanceoperators

Distance operators provide tools where distance plays a key role in the

analysis.

DISANCE calculates the true Euclidean distance o each cell to the

nearest o a set o target cells as specied in a separate image.

SPDIS is the equivalent o the DISANCE module, except that it

accommodates the special case o spherical distance units (degrees,

radians).

COS calculates a distance/proximity surace where distance is

measured as the least cost distance in moving over a riction surace.BUFFER creates buers around any set o specied eatures in an

image.

Te next set o our modules is used when rictions act with dierent

strengths depending on the direction o movement.

VARCOS computes an anisotropic cost surace using direction and

magnitude orce images.

DISPERSE models movement caused by anisotropic orces in terms

o direction and magnitude but unlike the VARCOS module, these

terms have no motive orce o their own.

RESULAN computes the resultant orce vector (as a magnitude

and direction image pair) rom two input orce vector image pairs.

DECOMP decomposes a orce vector (as a magnitude and direction

image pair) into X and Y component images, or takes X and Ycomponent images and produces a orce vector image pair.

PAHWAY calculates the route o least cost distance between one or

more points and the lowest point or points on an accumulated cost

distance surace.

ALLOCAE perorms spatial allocation based on a distance or cost

distance image.

RELOCAE moves eatures in an image to a target set o eatures in

another image based on minimum distance.

HIESSEN produces Tiessen polygons around a set o irregularly

distributed points.

contextoperators

Context operators assign new values in an image based on the values

o its surrounding neighbors.

SURFACE calculates either the slope, aspect, or an analytical

hillshading model o surace cells rom a given input image o terrain

heights.

FILER applies 3 by 3, 5 by 5, 7 by 7, or user-dened kernels to

calculate new values based on neighboring values. Te ollowing

lters are available: mean, Gaussian, median, standard deviation,

adaptive box, mode, Laplacian edge enhancement, high pass, Sobel

edge detection, and user-dened.

PAERN computes various numerical pattern indices (relative

richness, diversity, dominance, requency, ragmentation, and others),

using a 3 by 3, 5 by 5, or 7 by 7 template.

EXURE calculates measures o variability (ractional dimension,

class requency, edge analysis, and others) using a 3 by 3, 5 by 5, or 7

by 7 template.

GROUP identies unique contiguous polygon areas in an image.

VIEWSHED determines all visible cells rom any single or multiple

location on a surace or the proportion o viewpoint cells rom which

a viewshed is visible.

WAERSHED calculates all cells belonging to the watersheds o one

or more target cells.

HINERLAND determines the supply area dominated by point

demand centers.

PIXEL LOCAION creates new images representing the X and Y

coordinate o each cell center.

IDRISI includes tools or the derivation o runo potential and stream networks.


3/12

statistics

Te Statistics submenu provides a series o tools or perorming both

traditional statistical analysis and specialized spatial statistics routines.

HISO provides a graphic or numeric requency histogram and

statistics o the cell values within an image.

EXRAC calculates summary statistics or a set o input maps.

PAERN computes various numerical pattern indices (relative

richness, diversity, dominance, requency, ragmentation, and others),

using a 3 by 3, 5 by 5, or 7 by 7 template.

COUN calculates a relative requency probability image derived

rom a set o input Boolean images.

REGRESS undertakes a linear regression analysis with summary

statistics and graphs on image pairs.

MULIREG perorms a multivariate regression analysis between

images, one dependent variable and two or more independent

variables.

LOGISICREG perorms a logistical regression analysis on images,

one dependent variable and two or more independent variables.

MULILOGISICREG undertakes a multinomial logisticalregression on images where the dependent variable is multi-

categorical.

REND calculates up to a 9th-order best-t trend surace between

pixel values and their positions within the image.

AUOCORR calculates the rst-lag autocorrelation coecient, using

a rooks case or a kings case o an image using Morans I statistic.

QUADRA perorms quadrat analysis, the character o a point sets

pattern, in terms o its variance/mean ratio or density.

CENER calculates the mean center (center o gravity) and

standard radius or a set o points.

CRAIO measures the compactness ratio o dened polygons.

CROSSAB perorms a crosstabulation or a crosscorrelation between

two qualitative maps.

VALIDAE calculates specialized Kappa measures that discriminate

between errors o quantity and errors o location between two

qualitative maps.

ROC calculates the Relative Operating Characteristic providing a

measure o the correspondence between a quantitative modeled image

showing the likelihood that a particular class exists.

SAMPLE creates systematic, random, and stratied random point

sampling schemes.

RANDOM creates a new image o specied dimensions withrandom values that obey either a rectilinear, normal, or lognormal

distribution, according to a user-specied mean and standard

deviation.

SANDARD converts the values in an image to standard scores.

SPLUSIDRIS and SAIDRIS import and export images and data

between IDRISI and S-PLUS and Statistica soware respectively.

decisionsupport

Te modules in this menu are unique in that they specically address

multi-objective, multi-criteria resource allocation decision problems,

as well as problems o assessing and incorporating uncertainty in the

decision making process.

Te Decision Wizard is an automated assistant that steps you through

the decision support process in IDRISI. Te Wizard acilitates the use

o WEIGH, MCE, RANK and MOLA.

WEIGH employs the Analytical Hierarchy Process to compute a

best-t set o weights through a pairwise comparison o actors in a

multi-criteria evaluation.

MCE perorms a multi-criteria evaluation by means o either a

Boolean analysis, Weighted Linear Combination (WLC) or Ordered

Weighted Averaging (OWA) o actor images.

RANK orders every cell in a raster image.MOLA perorms a multi-objective land allocation analysis using a

decision heuristic to resolve conficts.

SANDARD converts an image to standard scores.

FUZZY evaluates the uzzy set membership values (possibilities) o

data cells based on any o three membership unctions: sigmoidal,

j-shaped, and linear, or through a user-dened membership.

COUN calculates a relative requency probability image derived

rom a set o input Boolean images.

IDRISI has tools or the validation o classifed maps to reality including crosstabulation

and error matrix analysis.


4/12

MDCHOICE resolves conficts between competing objectives by

means o a multiple ideal-point procedure.

Te remaining modules in this submenu are used in the evaluation

and handling o error in geographic analysis.

PCLASS evaluates the probability with which data cells exceed or are

exceeded by a specied threshold based on the stated RMS error or

the input map.

BAYES evaluates the probability that an entity belongs to any o a

number o dierent sets.

Beliefemploys the Dempster-Shaer Weight-o-Evidence procedure to

evaluate the degree to which evidence provides concrete support or

a hypothesis (belie) and the degree to which that evidence does not

reute the hypothesis (plausibility).

RANDOM creates random images according to rectilinear, normal or

log-normal models.SAMPLE creates systematic, random, and stratied random point

sampling schemes.

ERRMA produces an error matrix analysis o categorical map

data compared to ground truth inormation and tabulates errors o

omission and commission, marginal and total errors, per-category

Kappa Index o Agreement, and selected condence intervals.

change/timeseries

Change and time series analysis tools identiy and quantiy change,

as well as predict the eects o change on the environment, at scales

ranging rom local to global.

IMAGEDIFF compares two quantitative images o the same variable

or dierent dates.

IMAGERAIO compares two quantitative images o the same

variable or dierent dates through ratioing.

CVA (Change Vector Analysis) compares two-band sets o images or

two dates and calculates the magnitude and direction o change.

CALIBRAE adjusts the overall numeric characteristics o an image

to match an external standard using either image regression, user-

dened oset and gain, or user-dened mean and standard deviation.

CROSSAB perorms a crosstabulation or a crosscorrelation betweentwo qualitative maps.

o analyze change over multiple dates, the ollowing modules may

be used.

PROFILE creates proles over space or over time.

SA perorms a standardized Principal Components Analysis or

time series data.

FA perorms temporal Fourier analysis o time series images.

CORRELAE calculates the Pearson Product Moment Coecient o

Correlation between a set o values in an attribute values le and the

values through a time series o images or each pixel o an image.

KENDALL calculates the monotonic trend in data over time using the

non-parametric Mann Kendall statistic.

KENDAL AU calculates a non-parametric statistic to estimate the

degree o correspondence between two ordinal level variables.

Media Viewer is a presentation utility that can play Windows video

(AVI) les and can create AVI video les rom a sequence o IDRISI

images.

SAS computes temporal statistics on a per-pixel basis across a

raster group o images.

COR produces the correlations o a spatial pattern between a single

image and each image in a time series.

Te ollowing modules are used in modeling uture change.

MARKOV analyzes two qualitative landcover images rom dierent

dates and produces a transition matrix, a transition areas matrix, and

a set o conditional probability images.

SCHOICE creates a stochastic landcover map by evaluating the

conditional probabilities that each landcover can exist at each pixel

location against a rectilinear random distribution o probabilities.

DISAGGREGAE redistributes the conditional probabilities o a

particular landcover type according to a designated pattern.

IDRISI includes an unparalleled suite o tools or multi-objective/multi-criteria decision

support. Multiple suitability problems are defned in terms o actors, constraints,

actor weights and risk-taking strategy. Then the composite suitability images or each

objective are used together to best allocate areas to each objective. Rather than one-

shot black-box solutions, this approach provides decision makers with understandable

and deensible methods that can be iteratively improved. Here, industrial and commercial

suitability are modeled.


5/12

NORMALIZE linearly adjusts the values or a set o quantitative

images so the values sum to 1.0 at each pixel.

LOGISICREG perorms a logistical regression analysis on images,

one dependent variable and two or more independent variables.

CELLAOM perorms a cellular automata set o operations according

to a set o rules or changing states.

CA_MARKOV is a combined cellular automata / Markov change

landcover prediction procedure that adds an element o spatialcontiguity as well as knowledge o the likely spatial distribution o

transitions to Markov change analysis.

GEOMOD is a landuse change simulation model that predicts,

orward or backward, the locations o grid cells that change over time.

VALIDAE calculates specialized Kappa measures that discriminate

between errors o quantity and errors o location between two

qualitative maps.

ROC calculates the Relative Operating Characteristic providing a

measure o the correspondence between a quantitative modeled image

showing the likelihood that a particular class exists.

surfaceanalysis

Interpolation

INERPOL interpolates a distance-weighted average or a potential

model surace given an input set o points.

INERCON interpolates a surace rom a set o digitized contour lines.

IN creates a constrained or non-constrained triangulated irregular

network rom isoline or point data.

INSURF interpolates a ull raster surace rom a IN model and the

original point attribute data.

GENERALIZAION creates a point vector le rom the vertices o an

input line le or thins vector point data according to a user-denedradial search distance.

LINOPN extracts the vertices o a vector line data le into a vector

point data le.

INPREP adds or removes points along an isoline given a user-

specied tolerance distance.

Spatial Dependence Modeler* provides a wide range o tools to learn

about the patterns o spatial dependence in a sample data set.

Model Fitting* allows the user to dene mathematical models to

describe the covariance relationships among sample data.

Kriging and Simulation* creates ull raster suraces rom sample dataand the models developed through spatial dependence modeler and

model tting.

HIESSEN produces Tiessen polygons around a set o irregularly

distributed points.

REND calculates up to a 9th-order best t trend surace between

pixel values and their positions within the image.

Topographic Variables

SURFACE calculates either the slope, aspect, or an analytical hill-

shading model o surace cells rom a given input image o terrain

heights.

Markov Chain Analysis, a technique or predictive change modeling, is supported in

IDRISI with the module MARKOV. Predictions o uture change are based on changes

that have occurred in the past.

* Tese modules access a modied version o Gstat.

IDRISI includes a wide range o surace generation, interpolation and analysis routines

Upper let, Triangulated Irregular Network model created rom digital contour data

Upper right, digital elevation model (DEM) created rom the TIN with original contour

overlaid. Lower right, illuminated DEM emphasizes relie. Lower let, alse colo

composite image (TM bands 234) draped over the DEM. All data in this screenshot are

rom the Sierra de Gredos area o Spain.


6/12

CURVAURE calculates the maximum rate o change o a curve

t through a pixel in both the direction o aspect and also in the

direction orthogonal to aspect.

FRACAL calculates the ractal dimension o a surace using a

3 by 3 neighborhood.

Feature Extraction

CONOUR creates vector isolines at specied contour intervals roma continuous surace.

OPOSHAPE classies a surace into eleven dierent eatures: peak,

ridge, saddle, fat, ravine, pit, convex hillside, saddle hillside, slope

hillside, concave hillside, and infection hillside.

PI REMOVAL creates an adjusted depressionless DEM in which

the cells contained in depressions are raised to the lowest elevation

value on the rim o the depression.

RUNOFF calculates the accumulation o rainall units per pixel as i

one unit o rainall was dropped on every location.

FLOW calculates the fow direction rom each pixel into its next

downhill neighbor.

RUSLE (Revised Universal Soil Loss Equation) simulates armland

and rangeland nonchannelized soil loss by water.

WAERSHED calculates all cells belonging to the watersheds o one

or more target cells.

SLOPELENGH calculates the longest slope length in a given raster

region.

FACE produces an image o homogeneity.

SEDIMENAION evaluates the net soil movement (erosion or

deposition) within patches, elds, or river basins.

ModelingTe diverse items in the Modeling menu unleash the power o rasteranalysis in IDRISI. Most modules in this menu are also located in

other areas within the menu structure. Several modules are only

ound here.

LCM(Land Change Modeler for Ecological Sustainability) is an

integrated soware environment or analyzing landcover change,

projecting its course into the uture, and assessing its implications or

habitat and biodiversity change.

EM(Earth Trends Modeler) is an integrated soware environment

or the display, manipulation and analysis o time series data.

Macro Modeler provides a very mature graphical modeling interacethat exposes all o IDRISIs GIS modules as objects that can be linked,

dynamically and with eedbacks, with map layers in an algorithmic

chain.

COM and PYHON menu items allow or the development o

stand-alone modules as add-ons to IDRISI using a scripting language

such as Python or a ull programming language such as C++, Delphi

or Visual Basic. Using COM, client applications can be written that

control all aspects o IDRISIs operations.

Image ProcessingAlongside the geographic analytical operators ound in IDRISI, the

Image Processing capabilities round out a ull suite o tools or theprocessing o spatial data.

restoration

RESAMPLE perorms a local ane transormation or the geometric

restoration o images and can be used to georegister an image to a

reerence system or to another le.

The Land Change Modeler or Ecological Sustainability is organized around a set o

tasks or landcover change assessment, change prediction, assessment o its impacts on

habitat and biodiversity, and the exploration o planning interventions.

The IDRISI Macro Modeler provides a graphic environment or the construction o

models. Facilities are included or batch processing o multiple images through the same

model and or iterative processing with the output o one iteration becoming an input

to the next.


7/12

LOCALAFFINE is used to rectiy images that have an embedded grid

o control points with precise known locations.

MOSAIC automates color balancing when adjacent overlapping

images are joined into a single larger image.

DESRIPE removes the striping caused by variable detector output in

scanned imagery.

RADIANCE converts raw Landsat data values to calibrated radiance

using lookup tables o gain and oset values.

AMOSC corrects remotely sensed images or atmospheric eects

using either the Dark Object Subtraction model, Chavezs Cos(t)

model, the ull radiative transer equation model, or the Apparent

Refectance Model (ARM).

NDVICOMP creates temporal composite images o NDVI imagery

using the maximum value or a quadratic mean.

SCREEN uses spatial autocorrelation to screen a hyperspectral series

o images or the presence o signicant atmospheric noise.

enhancement

SRECH increases the contrast in an image or the enhancement o

visual interpretation.

COMPOSIE produces a 24-bit color composite image rom threebands o imagery.

FILER applies 3 by 3, 5 by 5, 7 by 7, or user-dened kernels to

calculates new values using a mathematical operation on the original

cell value and its neighbors. Te ollowing lters are available: mean,

Gaussian, median, standard deviation, adaptive box, mode, Laplacian

edge enhancement, high pass, Sobel edge detection, and user-dened.

PANSHARPEN perorms a panchromatic merge using color space

transormation, principal component transormation, and local

regression transormation techniques.

transformation

PCA provides both standardized and unstandardized principal

components analysis.

CCA perorms a canonical components analysis transormation.

MNF (minimum noise raction) maximizes the signal to noise ratio

or a set o images.

SA perorms a standardized principal components analysis or time

series data.

FA (temporal Fourier analysis) perorms harmonic analysis on

temporal images.

COLSPACE perorms Hue/Lightness/Saturation (HLS) to Red/Green/

Blue (RGB) color space transormations.

EXURE calculates measures o variability (ractional dimension,

class requency, edge analysis, and others), using a 3 by 3, 5 by 5, or 7by 7 template.

HERMAL converts Landsat M Band 6 raw data values to

blackbody temperatures.

VEGINDEX calculates 19 slope-based and distance-based vegetation

indices rom remotely sensed images.

ASSCAP perorms the asseled Cap transormation.

fourieranalysis

FOURIER allows or the transormation o images rom the spatial

domain to the requency domain and back again.

ZEROPAD is used to prepare images used in FOURIER.

FILERFQ, FREQDIS and DRAWFIL all acilitate the creation

o lters to be applied to requency domain images to enhance,

suppress or remove particular requencies prior to perorming a

reverse Fourier ransorm. FILERFQ oers 26 types o lters.

FREQDIS creates a requency distance image that may then be

manipulated with RECLASS or FUZZY. DRAWFIL provides an

interactive display utility in which the user may use the cursor to

trace particular requencies to be masked out.

signaturedevelopment

MAKESIG creates statistical signature les or each inormationaltraining site class.

Endsig is used to create end-member (i.e., pure) signatures or use

with UNMIX.

FUZSIG produces signatures rom data that are assumed to be

inherently uzzy or ambiguous in character.

PURIFY perorms a parametric (Mahalanobis distance) or a

nonparametric (unsupervised clustering) purication on existing

training site data.

Earth Trends Modeler includes a coordinated suite o data mining tools or the extraction

o trends and underlying determinants o variability.


8/12

HYPERSIG creates statistical signatures rom hyperspectral data,

either rom training site data or rom spectral curve library les.

HYPERAUOSIG automatically develops signatures or hyperspectral

image data based on the Linear Spectral Unmixing logic.

SIGCOMP graphically displays and compares signatures.

SEPSIG provides statistical measures on the separability o signatures

over a given set o bands.

SCAER creates a scattergram o the band space between images

used in the creation o signatures.

hardclassifiersPIPED is a parallelepiped classier.

MINDIS is a minimum distance to means classier.

MAXLIKE is a maximum likelihood classier with options to speciy

prior probabilities as values or images.

FISHER provides image classication based on linear

discriminant analysis.

KNN is a k-nearest neighbor classier.

CLUSER perorms an unsupervised classication using a variant

o the histogram peak technique.

ISOCLUS is an iterative sel-organizing cluster analysis procedureusing a predetermined number o clusters.

ISODAA provides an unsupervised classication o input images

using an iterative sel-organizing data analysis technique.

KMEANS classies according to the K-means clustering technique.

MAXSE is a hard classier that assigns to each pixel the class with

the greatest degree o commitment based on a ull Dempster-Shaer

hierarchy describing all classes and their hierarchical combination.

MLP undertakes the classication o remotely sensed imagery through

the articial neural network multi-layer perceptron technique.

SOM undertakes either a supervised or unsupervised classication o

remotely sensed imagery through the articial neural network Sel-Organizing Map technique.

Fuzzy ARMAP undertakes either a supervised or unsupervised

classication o remotely sensed imagery through the articial neural

network Fuzzy ARMAP technique.

CA undertakes the classication o remotely sensed imagery

through Classication ree Analysis with automatic and manual

pruning options.

softclassifiers/mixtureanalysis

BAYCLASS employs Bayesian probability theory to express the degree

o membership o a pixel to any class.

MAHALCLASS calculates Mahalanobis distance to produce a new set

o signature classes.

BELCLASS employs Dempster-Shaer theory to express the degree o

membership o a pixel to any class.

FUZCLASS employs Fuzzy Set theory to express the degree o

membership o a pixel to any class.

KNN is a k-nearest neighbor classier that can express or each

category its proportion among the k-nearest neighbors.

MLP undertakes the classication o remotely sensed imagery through

the articial neural network multi-layer perceptron technique with an

option to output so activation level layers or each class.

IDRISI allows on-screen digitizing o training sites and the development and analysis o

signatures including signature comparison, measures o separability, and scattergrams.

Linear spectral unmixing is available in the module UNMIX. Three options are provided:

the standard unmixing method, a probability guided method, and an exhaustive search

method. The latter two methods allow the user to evaluate more classes than the

standard unmixing method.


9/12

SOM undertakes either a supervised or unsupervised classication o

remotely sensed imagery through the articial neural network Sel-

Organizing Map technique with an option to output so typicalities

or commitment layers or each class.

UNMIX is used to classiy remotely-sensed images using Linear

Spectral Unmixing (LSUalso called Linear Mixture Modeling).

HYPERUSP provides unsupervised classication or hyperspectralimage data.

HYPEROSP provides or hyperspectral image classication through

an orthogonal subspace projection approach.

HYPERUNMIX extends the capabilities o Linear Spectral Unmixing

to hyperspectral data sets.

HYPERABSORB provides or hyperspectral image classication

based on library spectra and continuum removal o absorption areas

and the correlation o these areas in terms o t and depth between

the library spectrum and the spectra rom an imaging data set.

BELCALC calculates the degree o membership that each pixel

exhibits or each o the classes or which training data has beenprovided using the logic o Dempster-Shaer theory.

Beliefperorms a Dempster-Shaer Weight-o-Evidence classication

and extends the logic o mixture analysis, allowing or the ability to

combine new evidence with existing knowledge.

HARDEN produces hard decision images rom the so classier

outputs o BAYCLASS, UNMIX, FUZCLASS, BELCLASS, or

MAHALCLASS by choosing the class that has the maximum value.

SEGMENTATION CLASSIFIERS

Tree tools available or classication rom image segments.

SEGMENAION groups adjacent pixels into image segments

according to their spectral similarity.

SEGRAIN is an interactive training site and signature

development tool or use with the segmentation results created rom

SEGMENAION.

SEGCLASS is a majority rule classier based on the majority class

within a segment.

hyperspectralimageanalysis

HYPERSIG extends the logic o signature development to the

special case o hyperspectral data. HYPERSIG creates and displays

hyperspectral signatures either rom training site data or rom

spectral curve library les.

ASDIDRISI imports the spectrometer data collected using the

Analytical Spectral Device (ASD).

HYPERAUOSIG automatically develops signatures or hyperspectral

image data based on the Linear Spectral Unmixing logic.

SCREEN uses spatial autocorrelation to screen a hyperspectral series

o images or the presence o signicant atmospheric noise.

HYPERSAM is a spectral angle mapper hard classier or

hyperspectral data using a minimum-angle procedure.

HYPERMIN is a minimum-distance hyperspectral hard classier

specically intended or use with image-based signatures developedusing training sites.

HYPERUSP provides unsupervised classication or hyperspectral

image data.

HYPEROSP provides or hyperspectral image classication through

an orthogonal subspace projection approach.

HYPERUNMIX extends the capabilities o Linear Spectral Unmixing

to hyperspectral data sets.

HYPERABSORB provides or hyperspectral image classication

based on library spectra and continuum removal o absorption areas

and the correlation o these areas in terms o t and depth between

the library spectrum and the spectra rom an imaging data set.

accuracyassessment

SAMPLE creates systematic, random, and stratied random point

sampling schemes.

ERRMA produces an error matrix analysis o categorical map data

compared to ground truth inormation.

Techniques or hyperspectral image analysis are available in IDRISI. Tools include

absorption spectra analysis using continuum removal or estimation o the degree o

support or members o a library o spectral response curves developed in a laboratory

setting, an unsupervised classifer, and several supervised classifers including orthogonal

subspace projection and linear spectral unmixing.


10/12

ReormatIDRISI includes tools that allow you to change the data and le type

o a le, reorient an image or vector le, change the extent o the study

area, change resolution, generalize the level o detail in the le, join

les together, and convert les rom raster to vector and vice versa.

CONVER changes the data type or le type o an image or vector le.

PROJEC reprojects the reerence system coordinates o image or

vector les.

RESAMPLE perorms a local ane transormation or the geometricrestoration o images and can be used to georegister an image to a

reerence system or to another le.

WINDOW extracts a rectangular sub-area o a larger image to create

a new smaller image.

EXPAND alters the resolution o raster images through pixel

duplication.

CONRAC alters the resolution o raster images through pixel

thinning or by pixel aggregation.

CONCA concatenates multiple images or multiple vector les into a

single image or vector le.

RANSPOSE rotates an image by 90 degrees in either direction and

can reverse the order o rows or columns.

MEAUPDAE updates the documentation les or all les in a

raster or vector group le.

RASERVECOR converts data between raster and vector ormats.

GENERALIZAION is used to generalize vector point and line data.

It can also generalize raster data by merging smaller regions into

neighboring regions based on a given threshold.

LINOPN extracts the vertices o a vector line data le into a vector

point data le.

Data EntryIDRISI oers a host o tools to acilitate data entry. In addition to

the data entry modules in this menu, conversion utilities or existing

data that are in non-IDRISI ormats are also included (see Import

and Export).

CartaLinxis a ull vector topological editor and spatial database

development tool also developed and distributed by Clark Labs. It

provides tablet as well as on-screen digitizing capabilities and a wide

range o data editing tools.

Edit is the IDRISI text editor utility or creating a variety o ASCII

related IDRISI ormat les.

ASSIGN assigns new values to an image.

INIIAL creates an image containing a single value.

UPDAE assigns single values to specic cells or rectangular groups

o cells.

UMRefacilitates the creation o reerence system parameter les

based on the Universal ransverse Mercator system or subsequent

use with PROJEC.

Database Workshop is a relational database manager, and lies at the

heart o IDRISIs support or layer collections that link vector eature

denition les to database tables. Database Workshop provides the

ability to create, edit and analyze database les in IDRISI. IDRISI uses

the Microso ADO and Access Jet Engines as the basis or DatabaseWorkshop. With this acility, one can undertake a wide variety o

database operations including queries, calculations, and map display.

Both the Calculate and Filter operations are supported through the

use o Structured Query Language (SQL).

IDRISI Explorer is a general purpose utility to manage and explore

IDRISI les and projects. Use IDRISI Explorer to set your project

environment, manage your group les, review metadata, display les,

and simply organize your data with such tools as copy, delete, rename,

and move commands.

Import and ExportIDRISI includes a variety o utilities or le import and exportbothgeneral conversion routines as well as modules or specic soware or

data ormats.

generalconversiontools

GDALIDRISI is a ront-end utility that interaces with the open

source GDAL raster translation soware.

GENERICRASER is an all-purpose utility to import raster data in

a variety o data types and ormats, including byte, integer and real,

Satellite imagery is easily imported by means o special import modules (e.g., Landsat),

industry standard ormats (e.g., GEOTIF) or generic import routines.

0


11/12

band-interleaved by line (BIL), band-interleaved by pixel (BIP) and

band sequential (BSQ) ormats.

CRLF adds or removes carriage returns or line eeds.

XYZIDRIS is used to import ASCII X,Y,Z coordinate data to a point

vector le such as might be collected by a GPS unit or might be

entered by hand into a spreadsheet or text le.

VAR2FIX changes variable-length ASCII les to xed-length les.

SSIDRIS is used to import spreadsheet data when the cells o the

spreadsheet are to be interpreted as cells in the resulting image.

MODISQC creates the quality assurance science data sets or MODIS

data products.

government/dataproviderformats

Landsat EM or Landsat NLAPS, FAS, GEOIFF or HDF ormats.

SPO or SPO satellite data in GEOIFF, SPO Scene (CAP), or

GEOSPOSPOView ormats.

GEOIFF or generic GEOIFF/IFF les.

HDFEOS or HDF-EOS4 ormats including HDF 4 and HDF-EOS 4.

GACPIDRISI imports Global Aerosol Climatology Project data into

IDRISI.

NECDF imports NECDF data into IDRISI.

OLRIDRISI imports Outgoing Longwave Radiation into IDRISI.

PSDIDRISI imports Physical Science Division standard ormat into

IDRISI.

XYZMONHLY converts the University o Delawares Center or

Climatic Research ASCII monthly x, y, and multi-z data into IDRISI.

ASDIDRISI imports the spectrometer data collected using the

Analytical Spectral Device (ASD).

SACIDRIS or SAC-C satellite data rom Argentina.

RADARSA or RADARSA International data.

GPCIDRISI imports the International Satellite Cloud and

Climatology Projects Global Processing Center data into IDRISI.

GOODE2LL or Global AVHRR 10-day composite data rom USGS

NASA DAAC in the Goodes Homosoline projection.

SDS or Raster Spatial Data ranser Standard data.

DLG or Digital Line Graphs (Optional Format) data.

CG or the Composite Teme Grid data.DEMIDRIS or USGS Digital Elevation Models.

desktoppublishingformats

BMPIDRIS or Windows Bitmap les (BMP).

DXFIDRIS or CAD DXF les.

GEOIFF/IFF or agged Inormation File Format les (IFF).

JPGIDRIS or JPEG les.

software-specificformats

IDRISI provides ull map composition capabilities, including multiple raster and vector

layers, layer blending, interactive RGB compositing, multiple legends, title, text labels,

grid, north arrow, scale bar, text and image insets. This image depicts the White Mountain

region in New Hampshire, USA, including data or elevation, roads, and boundaries

using various layer blending options.

The Explorer bar provides the unctionality o the older IDRISI Explorer, MetaData

Data Paths and Collection Editor modules. The bar can be minimized against the

let-hand edge whenever more workspace is required. Shown here are the Files and

Metadata panels. Double clicking an image fle causes it to be automatically displayed

Similarly, images can be dragged and dropped into any IDRISI dialog. Also shown in

this illustration is the use o one o the two measure tools in this case, the zone

measurement tool. The linear measurement tool measures the length and azimuth o a

single or multi-segmented line. Note also the new grid ticks option and one o the new

north arrow styles.


12/12

To order

IDRISI: The Taiga Edition,

contact us at

www.clarklabs.orgor at

Clark Labs

Clark University

950 Main Street

Worcester, MA 01610, USA

Tel: +1.508.793.7526

Fax: +1.508.793.8842

Email: [email protected]

SHAPEIDR,ARCRASER, and ARCIDRIS or ESRI Shape

les, ArcIno Raster Exchange, and ArcIno GENERAE/UNGEN le

ormats.

ALIDRIS or Atlas*GIS BNA les.

ECWIDRIS or ECW les.

ENVIIDRIS or ENVI les.

ERDIDRIS or Erdas LAN and GIS les.

ERMIDRIS or ER Mapper les.

GRASSIDR or GRASS raster les.

MAPIDRIS or Map Analysis Package les.

MIFIDRIS or MapIno Interchange les.

SRFIDRIS or Surer GRD les.

SPLUSIDRIS or SPLUS statistical les.

SAIDRIS or SAISICA les.

DisplayDisplay and map composition utilities provide tools or visualization

and enhancement.

DISPLAY Launcher is the main display acility or raster and vector

IDRISI les.

ORHO is a acility that creates orthographic perspective (3-D)

displays o digital elevation models (DEMs) or any continuous

raster image.Fly Trough is an interactive 3-D viewer using OpenGL technology

that allows users to simulate movement through space using existing

IDRISI images.

Media Viewer is a presentation utility that can play Windows

video (AVI) les and can create AVI video les rom a sequence o

IDRISI images.

Symbol Workshop allows one to create and modiy symbol and palette

les or vector and raster display.

COMPOSIE produces a 24-bit color composite image rom three

bands o imagery.

SEPARAE perorms color separation o palette images into

RGB components.

ILLUMINAE is a hillshading merge acility.

HISO provides a requency histogram and statistics o the cell

values within an image, presented graphically or numerically.

SRECH increases the contrast in an image or the enhancement

o visual interpretation.

System Requirements- Intel Pentium IV or higher

- Windows XP and above

- Minimum display o 1024 x 768 with 64,000 colors

- 1 GB RAM or greater

- 1.3 GB hard disk space space or installation

- CD-ROM or installation

About Clark LabsClark Labs is dedicated to the research and development o geospatial

technologies or eective and responsible decision making orenvironmental management, sustainable resource development and

equitable resource allocation.

Based within the world-renowned Graduate School o Geography at

Clark University, Clark Labs is known or pioneering advancements

in areas such as decision support, uncertainty management,

classier development, change and time series analysis, and dynamic

modeling. Partnering with such organizations as the Gordon and

Betty Moore Foundation, Google.org, USDA, the United Nations and

Conservation International, Clark Labs leverages its academic base to

develop innovative and customized research tools, provide soware

solutions to organizations in need and apply geospatial expertise to a

range o real-world problems.

idrisi taiga gis image processing technical specifications

Documents