Leaderboard

Hello GIS forum ! Where can I find a complet tutorial about Land Use/Land Cover classification using LANDSAT images ? A tutorial for begginers, for newbies, non-experts in remote sensing ? I would like to find a tutorial that presents step-by-step, from A to Z, all the aspects of using LANDSAT images to obtain a land cover map. Content of the tutorial: Where to obtain and how to select satelite images of LANDSAT data collection ? (this I allready know: Glovis, EarthExplorer, and 0% clound cover, different seasons for vegetation status, etc) The pre-processing steps of the satellite images ? How to make an atmosferic calibration, remove any disturbances, etc ? What software to use better : ENVI or ERDAS ? What modules of these software to use for pre-processing ? Classification methods: un-supervised classification/ supervised classification ? (advantages/dis-advantages of these metods?) How to make a supervised classification (selecting the classes according to spectral values, selecting samples/ training areas, spectral mixture, spectral analysis, spectral plots/diagrams, etc) What is the best software to use for this step (ENVI or ERDAS) and what modules to use ? Post-classification methods ? How to make a post-classification of a Land use /Land cover image ? (in ENVI or ERDAS) Tips and tricks ? ( is it better to change the classification method, the samples, or the satelite image ?) What is an error matrix of the classification and how to make it ? How to asses the quality of the classification method ? Are there any methods to obtain good results in classification of LANDSAT images ? So this is what I am searching, a tutorial about using LANDSAT images to create a Land use/Land cover map ! ( video tutorial or document tutorial with a package of images and examples, explained step-by-step, special for begginers or "idiots" ) Please give me a tutorial !!!! Thank you all, my friends !

Anyone familiar with ThePirateBay, biggest Torrent sites in the world ? okay, their hosting company, PRQ, the most controversial Hosting company (they host anykind of websites, regardless the content) is raided All sites hosted on the 80.88./19 net are currently unavailable, including but not limited to the torrent sites torrenthound.com, linkomanija.net and tankafetast.nu, release blog RLSLOG.net, and the sports streaming sites atdhenet.tv, hahasport.com, sportlemon.tv and stopstream.tv. Complete news : http://torrentfreak.com/prq-police-raid-takes-down-dozens-of-file-sharing-sites-121001/ Sad news for freedom of sharing

Remote Sensing is an advanced thing, so there is a good possibility that you'll never find anything 'complete'. It's not just software, you'll need to learn the theory as well. Riccardo Klinger has a good set of tutorial video you can try. http://www.youtube.com/playlist?list=PLCEBF53A31190DABE&feature=viewall Speaking about Erdas and ENVI, both works fine in lots of aspect. Sometimes ENVI runs smoother than Erdas, runs in linux too. The quality of result relies on the methodology and data you follow. From my pov, try object-based classification rather than pixel-based. Object based image analysis (OBIA) is one of the most powerful technique to use most from your image and give far better result than pixel based analysis. In OBIA you are working with color, texture, pixel statistics, nature,and band values all at once. Pixel based classification only rely on one thing-DN, so error is inevitable. It has an automated workflow, so no raster filling and reckoning fuzzy-class. Throwing my two cent.