Tag Info

Hot answers tagged

18

I would have to say that the most complete software environment for Machine Learning and nonparametric modeling is R. This is a big field in statistics, spanning K-NN, Kernel smoothing, General Additive Models, weak learners, support vectors, neural nets, semi-parametric spline regression, imputation, etc... I would highly recommend reading: Hastie, T., R. ...


17

There are a lot of factors influencing GPS accuracy. James Ryan is right about night time and clear atmospheric conditions. Then there is the satellite "layout" on a specific time at at specific position on earth. Depending on how many you can "see" (4 at least for a 3D fix) and their distribution (all in one row is bad, evenly spread out better) your ...


16

With GRASS GIS (http://grass.osgeo.org) you can do these steps: Import script, see page bottom at http://www.grassbook.org/examples_menu3rd.php i.landsat.rgb - auto-enhance colors i.landsat.toar (addon for GRASS 6, included in GRASS 7) - convert DN to top of atmosphere radiance i.atcorr - correct top of atmosphere to surface reflectance i.landsat.acca ...


15

I agree with @vascobnunes opinion but if you want to define certain objects you have to use LANDSAT TM because more classification needs more bands as (R, G, B, NIR, MIR, TIR, FIR)... and my choice is that you should use LANDSAT TM (I gave same information in the following explanation) for vegetation. The important thing in this case is that you should look ...


15

IDL is a fantastic stand-alone programming language (you do not need ENVI). I particularity like it for very fast matrix processing on large arrays. @Aaron makes IDL sound much less flexible then it really is. The majority of IDL development came out of the Physics and Astronomy communities. There is robust support for mathematical and statistical ...


13

I use a number of tools depending on the type of classification I am trying to perform. For general unsupervised/supervised classification I use ENVI, which has many options for classification methods (including some more advanced methods using neural networks and support vector machines). It is very easy to extend ENVI using the IDL programming language, ...


13

As Chad Cooper mentioned, what you want to perform is called Object-Based Image Analysis (OBIA). It's a fairly complex process which segments and then classifies an image. There are many programs out there which will perform this for you. However, you will require high-resolution, multi-spectral imagery. Incorporating LiDAR will probably help you out ...


13

From a remote sensing perspective, the main benefit of IDL is that it extends the capability of ENVI similar to how the Python arcpy site-package extends the functionality of ArcGIS. If you will not have access to the ENVI platform, consider learning a different programming language. Additionally IDL is a commercial product whereas Python is open-source and ...


12

The general approach for performing this kind of disaggregation is through dasymetric mapping, which uses ancillary data to inform the spatial distribution of a phenomenon, and is often used for population analyses, such as this one in San Francisco. This paper provides good background on the technique, and if you're working in ArcGIS, scripts have been ...


12

Great Question: Can Time and Position affect your GPS accuracy. Short Answer: YES Why? First lets place your GPS Receiver in perfect conditions where the atmospheric conditions are perfect, there are no multipath effects, no radio interference, and line of sight to GPS Satellites is clear. Your Time and Position on Earth will determine where the GPS ...


12

I am in Canada so if I need this imagery I can get it free at Geobase. Elsewhere you should be able to download from USGS direct. You will need to register on both sites. Here are NASA links to download free Landsat data.


12

I have used OpenCV in the past to train for object detection for geo. Orfeo Toolbox is a good open source choice as Vascobnunes pointed out. For a closed-source version, you can take a look at Feature Analyst (that also has an ArcGIS extension). At the end, it boils down to training a support vector machine. There are several libraries that you can use for ...


12

There are a few good ones around: Orfeo Toolbox GRASS SAGA GIS All with the bonus of being able to be used though the QGIS interface using the SEXTANTE plugin like so http://blog.orfeo-toolbox.org/uncategorized/otb-inside-sextante-inside-qgis


11

You can use GRASS GIS for this which supports texture extraction and image classification based on a radiometric/segmentation approach. For an idea, check this conference abstract, a planned talk at the Geoinformatics FCE CTU 2011. See also: http://grass.osgeo.org/wiki/Image_processing and http://grass.osgeo.org/wiki/Image_classification for an overview.


11

You can't 'remove' clouds from optical imagery, what you see is what you get; they are photographs and there is no optical data recorded from below the clouds in the same way that there is no data underneath building roofs. If you use remote sensing data of a longer wavelength than light such as microwave, the water particles in the clouds do not absorb the ...


10

If you only have SPOT 5 and Landsat TM to choose from, money is not a problem and for a small area of 30 000ha, I would agree that SPOT5 is the best choice, although Landsat would have some strong advantages: SPOT5: 2,5 m spatial resolution 3 spectral bands (Green, Red, Near Infra-red) about 2,64€ per sqkm for new acquisitions good revisit time biggest ...


9

I'd strongly recommend scikits-learn for Python. It supports supervised and unsupervised classification and the documentation is excellent (particularly check out the Machine Learning for Astronomical Data Analysis tutorial and the accompanying YouTube video (note: this is 3 hours long)). The project is under active development, with the last version being ...


8

An alias, in signal processing which is what we're dealing with when we are talking about images, is when a signal is sampled at a resolution that makes it impossible to recreate the original signal exactly. Take this 1-dimensional case: The original signal is the purple sine wave, and the blue dots are where it has been sampled. The blue line is the ...


8

Please mention the sensor of Landsat 5, is it MSS or TM? Assuming it is Thematic Mapper data, you have visible red and shortwave infrared data. You can directly infer from the band reflectance values about where vegetation patches lie and hence moisture content. Band 3 (Red) can help you discriminate vegetation slopes and Band 5 (SWIR) can help you ...


7

I am afraid satisfying roof detection cannot be achieved with only one single satellite image. You should try to use other sources of information. The following article describes a method using a DEM + aerial image pairs + cadastral data: M. Durupt, F. Taillandier. Automatic Building Reconstruction from a Digital Elevation Model and Cadastral Data: An ...


7

First problem: You're looking at a mixture of minima. One gigantic tree with an acre-sized crown looks quite a lot, interpreted on a point / kernel density basis, like a field with no trees at all. You will end up with high values only where there are small, rapidly growing trees, at edges and in gaps in the forest. The tricky bit is, these dense smaller ...


7

For processing of Landsat, I can recommend GRASS. I tried many others. You may need to refine your question with regard to the type of imagery you propose to use. There are workflows which have been more or less developed and implemented in various software. Not only the type of imagery, but the purpose of the processing and final analysis. For Landsat, ...


7

Erdas used to work together wih ESRI, but now it is ENVI that has joint its forces. I would therefore use ENVI for the compatibility. But if you are looking for an good open source solution, I recommend Orfeo Toolbox (http://orfeo-toolbox.org/otb/ ). You can either use the library, the command line application or a complete GUI (called Monteverdi). ...


6

The general process for solving this problem is by orthorectification, to convert the data into a Cartesian coordinate space, where each cell represents approximately the same spatial extent. The OSSIM package provides orthorectification, as does GRASS. You'll need ancillary data to perform the rectification, and it can be a time consuming process, but ...


6

Generally the answer is "Quite a large effect", but it can vary significantly. In general - do atmospheric correction if possible, and the best atmospheric correction you can - although I appreciate this isn't always easy. As for a few more details: Contributions from the atmosphere to NDVI are significant (McDonald et al., 1998) and can amount to ...


6

You could try the gdal_fillnodata tool which is also available in QGIS via the Raster->Analysis->Fill nodata menu. It uses an inverse distance weighting (IDW) interpolation. I just tried both that method and the single date Triangulation interpolation (in ENVI) and gdal_fillnodata looked much better. If you want to merge multiple dates, you might have to ...


6

The "best" software is somewhat subjective and dependent on your needs. All of the options provided thus far are worth exploring. I would like to add SPRING software to the current suggestions. This is a very robust free GUI-driven software for remote sensing. All of the functionality that you mentioned is available.


6

Generally speaking, if you have a Single Image (Single or Multi Band), you cannot get Elevation information from it directly. If you have a Stereo Image Pair, you might be able to get some Elevation values from it, but even those are not accurate without a rough base DEM, or Benchmark. You also get satellite data, which consists exclusively of a DEM. For ...


6

There are published coefficients available for MSS, TM5 ETM+7, QuickBird and IKONOS but I do not believe that anybody has derived coefficients for Rapid Eye. Here is a paper that describes how the authors derived the coefficients for Quickbird (http://www.asprs.org/a/publications/proceedings/pecora16/Yarbrough_L.pdf).


6

To better understand, I have done a similar segmentation in GRASS GIS 7 using the new i.segment. In my opinion these lines appear where the image data were mosaiked due to a non-perfect histogram matching (or whatever) being used. In short: Orfeo or any other segmentation software may deliver better results when the initial mosaiking is improved to avoid ...



Only top voted, non community-wiki answers of a minimum length are eligible