This model accesses the pixels in a 3 by 3 matrix around the center pixel. It just averages the nine pixel values, but you can add your own algorithm easily by modifying the function. It can be used for change detection in thematic imagery.

Submitted by Ross Lincolne.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

This model normalizes brightness trends across image frames. For any single band image, the model calculates a view zenith and a view azimuth value for each image pixel. Solar azimuth, solar zenith, and flight heading information are then used with the view zenith and view azimuth values to calculate a scattering angle value for each pixel in the image, hence creating a scattering angle image. This scattering angle image is then transformed into a thematic image containing 255 classes; each single class containing pixels of equal scattering angle value. For each scattering angle class, the mean and standard deviation of original image pixel values is calculated. This can be imagined as an overlay function where the mean and standard deviation of input pixels are calculated for pixels which fall within a given scattering angle class, or "band"; the calculation is then repeated over all scattering angle bands.

Submitted by:
Lloyd Coulter
Department of Geography
San Diego State University

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

When the values of a raster data set such as thematic images are to be written to the correspondingly located points in an ArcInfo point coverage, this model is very useful. A thematic raster such as a classified image and a point coverage are the inputs. Both the datasets should overlap. In addition a numeric item may be added to the point coverage. Raster values can be written to this item. The procedure is as follows:

1. Select the thematic raster file.
2. Next, select the ArcInfo point coverage to which the raster values are to be assigned. The point coverage should preferably have a blank item to which the raster values are to be written. The custom matrix has just one cell with value of 1. The output ArcInfo point coverage is the same as the input coverage.
3. Make sure to see that output button is clicked
4. Output to Descriptor or Attribute is checked
5. Output to a: Existing Layer
6. Select the input point coverage
7. Select Point Feature Type
8. Select the newly added Item. Once the model is run, the raster values corresponding to the points are written to the selected item in the point coverage.

Submitted by:
Samuel Rajasekhar
Minnesota Department of Natural Resources
Forestry Resource Assessment
Grand Rapids, Minnesota, USA

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

This model produces an output image with the ratio of band 2 to band 5. It is useful for detecting water vs. land, such as shorelines, etc.

Submitted by Ross Lincolne.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

These are two spatial model which can be used to calculate back-scatter (Sigma Nought) coefficient from radar imagery.

The input images are a DEM and a radar image (the model is set up now for IMAGINE's example data, but you can change it to any images you want). Two images are produced by the model, an incidence angle map and a cosine (inc. angle) map. Cosine (inc. angle) is the most common conversion to sigma-0 but many of people use difference trig functions so one can use whatever function he wants on the angle mask.

This is discussed in Bayer, Thomas, and Schreier, Gunter, "Terrain Influences in SAR Backscatter and Attempts to their Correction", IEEE Trans on Geo. and Rem. Sens., vol. 29. No. 3, 1991.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

Created by Paul Beaty.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

This is a matrix model which will create a N x M thematic file in a From/To matrix setup. It is useful in change detection matrix operations. In this case there are 15 classes in two dates of imagery, so the resulting image is a 225-class thematic file. Image Interpreter > Utilities > Summary will provide the same information.

Submitted by Eric Dobson.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of ERDAS software.

These models convert DEM elevation units from meters to feet, and feet to meters. The models also round the elevation-calculated unit to the nearest integer. The loss of precision during the rounding process does not significantly affect the accuracy of USGS DEMs. While loss of precision may be a cause of concern, rounding float data into 16-bit signed integer data allows the ERDAS IMAGINE RLE compression routine to achieve a good compression ratio. These models assume a null value of -32767. The value is ignored in the statistics calculation. If your null data values are different, simply change the "Ignore" value before running the model.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

This model rounds data to the nearest tenth, working around a limitation in the Modeler.

How this model works: Multiply the DEM values by 10, then round those values to the nearest integer, then finally divide those rounded integer values by 10, keeping float single (or better) throughout the model.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

Rather than printing directly to a printer, or creating a file in a particular printer format (for example, HP RTL), users sometimes have a requirement to create digital layer separates from an image or Map Composition. For example, separate digital files of the red, green and blue (RGB) components of a Map Composition might be required as input to a Film If the color separates required are RGB, then their creation is simple: The three bands of an image used for the RGB display can be exported to three generic binary layers; or a Map Composition can be printed to .IMG (which will create a three-band RGB image) and again these exported to the format accepted by the printing device.

However, as printers generally use ink instead of light to create a visual image, the primary colors of pigment - cyan, magenta, yellow - are used in printing, instead of the primary colors of light (red, green, blue). If the final print method uses cyan, magenta, yellow (CMY) separates, rather than RGB, then a longer methodology is required to create the separates.

The model rgb2cmyk.gmd can be used in the ERDAS IMAGINE Spatial Modeler to convert an RGB image to CMYK.

(ERDAS IMAGINE 8.3.1 and higher)

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

After reprojection, the warped image may have sliver areas of null data (zeros in IMAGINE). This model converts the IMAGINE null values (zero) to ESRI null value (-32767).

The simple model converts zeros to ESRI null, and is useful if you have already-reprojected images, without the original un-reprojected images.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

Concept

This model is used to calculate the brightness temperature from band 6 TM. Unlike the published TM-temperature model on the ERDAS web site, this model would consider the min-max digital count for every image. As we know, to get the spectral radiance, Lλ, the following algorithm (1) is applied,

Lλ= {(Lmax-Lmin)/DNmax }* DN +Lmin (1)

where,

• Lmax= the spectral radiance that is scaled to DNmax
• Lmin= the spectral radiance that is scaled to DNmin
• DNmax= Maximum DN
• DN = the quantized calibrated pixel value.

This is a linear scaling equation, which corresponds to (2)

Lλ= Slope *DN + Intercept (2)

The DNmax and DNmin for every image are virtually different as the land cover is different. However, the given Lmax and Lmin are fixed. Thus, a proper linear adjustment on Lmax and Lmin that are equivalent to particular scene DNmax and DNmin is necessarily.

Model

The Global Max and Global Min in the model are used to find the DNmax and DNmin for the input data, IGNORE is applied to Global Min to prevent zero value taken into account, normally zero comes from dark area after image was geocoded.

“Lmax adjustment” and “Lmin adjustment” are used to calculate Scene Lmax and Scene Lmin.

Submitted by:
Woo Shu Chuen
Researcher
Malaysian Centre for Remote Sensing (MACRES)

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

This model converts all zero values to ones. This can be useful, for example, when printing images with collar areas of zero values, to avoid wasting ink.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

For use with ERDAS IMAGINE 8.4+ 

This model converts a thematic color-mapped image, such as a scanned map, into the equivalent 3-band RGB image. This is very useful if you need to mosaic several thematic datasets when a consistent scanning color-map hasn’t been used (i.e., if in image 1, DN value 1 represents blue, whereas in image 2, DN value 1 represents green - the only way to successfully mosaic these files and retain the correct colors is to break the colors out into RGB bands). 

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.

These models convert the spectral radiance of Landsat 4 and Landsat 5, band 6, to temperature values in degrees Celsius. The models search for the sixth band in a Landsat data file and will not work unless at least six bands are present in the imagery. Furthermore, extraneous variables can not be fully accounted for, such as sensor inconsistencies or cloud cover. The output can be converted to degrees Fahrenheit by the formula 1.8 * ºC + 32. There are two download files, one for Landsat 4 and one for Landsat 5.

Disclaimer: This user-submitted download is provided for the benefit of all ERDAS software users, but is not supported by ERDAS, and may not be up-to-date for the latest versions of the software.