How to calculate NDVI with ArcMap 10.1

[@aestfd]

thanks man, you are a life savior, because I don't know how to use ENVI. I appreciate this a lot

[@sounny]

This is a great idea for my next video! What you need to do is use the reclassification tool. What you will do is go to the ArcToolbox and look in the spatial analysis toolbox for reclassify. The reclassification functions reclassify or change cell values to alternative values using a variety of methods. So what I suggest is that you reclass 0.7-0.9 to a value of 1, and the others to 0. Then with the result, look at the attribute table for a pixel count. Multiple pixel count by size for area

[@IAmTheRocketMan]

Easy to follow and understand. Thank you !

[@jackgreenhalgh7664]

Perfect pace and info - thanks :)

[@sounny]

I used to use ERDAS to do this, but I am glad I can do it with ArcMap now.

[@БексултанКанатов-ю3и]

Sipmly incredible tutorial!! Made of great deal out of it!! Many thanks to author!!

[@sounny]

Thanks! I am glad that it helped.

[@LezanKOthman]

Thank you soooooooo much... hope to see more videos from you...

[@nico1764]

Thanks for this video from Santiago of Chile!!..

[@sounny]

It is exactly the same product. I actually find it easier to calculate in ArcGIS. Before I used to make the NDVI in ERDAS and then bring it into ArcMap. This saves a step. ENVI also uses the same formula.

[@mohamedhassan-dx7ko]

the video was nice i wonder wich one is more accurate and reliable to work with the Red Edge Normalized Difference Vegetation Index or the normal NDVI am kind of stuck so i would really appreciate ur help

[@leobuffon9374]

Hi there. It looks great this could be done in ArcGIS. I am wondering if there is any concern about the sensor selection step (e.g. Landsat TM) in Erdas/ENVI? My problem is cannot find the matched sensor for DOQQ data! Thanks again for this great vedio

[@surajdixit]

Thanks for this video, its very informative!!

[@aestfd]

How is your experience about it, is it too different the final product you got? I mean in the quality of it.

[@LezanKOthman]

Thank you for this intersting vid. I wonder how can I calculate the total area of healthy vegetation? ranged between for e.g (0.7-0.9)? Any Ideas?

[@sounny]

Nearly all satellite Vegetation Indices employ this difference formula to quantify the density of plant growth on the Earth - near-infrared radiation minus visible radiation divided by near-infrared radiation plus visible radiation. The result of this formula is called the Normalized Difference Vegetation Index (NDVI). If you can tell me what two formulas you are comparing perhaps we can figure out which is the best.

[@knallhardt]

hey there, thank you for the lesson. its a nice explaination. is it possible to identify different kind of crops like for example potato, corn?

[@prodanmn]

good work, thx for this lesson

[@dr.zeyadal-saedi2012]

thanks for this lesson, I have two main question, first you should used band 4- band 3 not band 1 as you mention second you used image with extension Jpg and you should used img extension because Jpg is a view without coordinates thank you

[@j.r.p.jespersen8860]

Great tutorial, although I can't get mine to calculate the raster properly. ArcGis don't give much of an explanation as to why. You got any idea, what might cause the calculation to malfunction?

[@sounny]

Glad you enjoyed.

[@08drano]

I need to combine four quadrants to for my study area. Should I mosaic them first and then do analysis, or do I need to do them individually and then mosaic. I tried mosaic them together first, but then the bands were no longer giving me an NDVI.

[@kosmikastaway1285]

Excellent tutorial! As a newcomer to NDVI and ArcGIS in general I was able to get a decent NDVI image! Thank you! I was wondering how you might select an area from your image and get an average NDVI for that area?

[@sounny]

If you crop your study area, and then look at the layer properties under the source tab you will see the statistics. There you can find an average value.

[@kosmikastaway1285]

Thank you! I will give it a go!

[@abeeyabiodun7717]

Please I need to extract the number of buildings in an urban environment am working on for my research work. Is there a way to get this done without having to count manually.

[@aeestay]

This video is great thank you! one question, do you know where can I get maps either aerial or not for Mexico? I`m trying to make a NDVI for the volcanoes area. Thanks!

[@neerajdhingra8216]

Hi, I would like to ask you, what if we have data with 3 bands instead of 4 bands. Could you please show how to work in case of 3 bands or could you share a link if you have already shown before. Thanks!

[@maksimfesenko2255]

Can you make a video on how to use 3 bands ? because i downloaded maps from earthexplorer & glovis , and they all have 3 bands

[@sounny]

It is more important to know what is the IR and Red bands, then you will make the idex off those two. If the image is just RGB then you can't really make a NDVI.

[@sounny]

@mohamed hassan You will need to look into model builder for that.

[@Archil8884]

Well, I have Landsat 4-5 TM tiffs with 7 bands. I`m making one tiff with with 6 bands-1st, 2nd, 3rd, 4-th, 5th and 7th bands (6th bands resolution is different and isn`t suitable for my task) , from here band 3 is red and band 4 is NIR and I`m doing everything accordingly of this video, but after map algebra, I get just black quadrangle, nothing more. How I can solve this problem?

[@pedrovieiracandido9317]

The NDVI formula considers reflectance values, so in theory you would have to convert the Digital Number (DN) to radiance then to reflectance before adding and subtracting the band values...

[@pedrovieiracandido9317]

i am not sure how you would obtain the metadata used in the conversion for the photos though, with the Landsat and RapidEye programs it is rather easy.

[@sounny]

Pedro Vieira Yeah you are right about the DN, but these are aerial photos. The one thing is that the index is unitless, so as long as the dn to radiance conversion originally is linear then there is no problem using the dn numbers instead of radiance because the units cancel out. With the Landsat and Rapid eye products you don't have as high of a resolution but they are easier to convert these numbers.

[@jou123ish]

Thanks for this tutorial - I successfully created three files with NDVIs. Two files had NDVIs ranging from -1 to 1, and one had an NDVI ranging from -1 to 0.7. I am trying to average the NDVI values for specific geographic areas, and am having a hard time figuring out what format these files are on. Are they grid files? Raster files? My computer doesn't recognize them as raster datasets.

[@sounny]

The one I made in the video is a grid. When there is no extension on the name ArcMap makes it in a ESRI Grid format. It only really works with GIS platforms. If you want, you could add a .tif to the end of the name so it will be in the uncompressed raster format that windows can open as well. Be careful when using compressed extensions like Jpeg, because there will be some pixel value averaging, and you really care about the individual values of each pixel.

[@taufiknuradikusuma6337]

thank for the tutorial...

[@essam8697]

Thanks for sharing this video. I use Rapideye for my imagery and I know that all vegetation indices use "Reflectance" in the NIR/Visible range to get a value. However, these bands and their values that you are using seem to show digital numbers DN (not the reflectance); this is the case for Rapideye imagery that I have just received. Therefore, one needs to convert the DN to the reflectance first before calculating NDVI for instance. Please let me know what you think. Thanks

[@sounny]

This is a very good question and I wish I could give you a great answer. But I will give you the best I can. Hoefully if you find something better you can post it to here. So basically in this video we were using aerial photography which had an infared band. This value isn't really a reflectance value, in that the values are unique for each shot. In your case you are using rapideye imagery so there is a better chance of getting reflectance numbers. The digtal numbers in this video here are a range from 0-255, or basically 8 bits. It is a standard RBG image. So the creators of the image may have used some kind of histogram equlaization. If they did it can be a problem. But perhaps they did not. The only thing that still gives hope to these values is that NDVIs are unitless. So the unit would cancel anyways and should result in simillar values as if you did it with relectance vale. Hope that helps!

[@80Hgmercury]

Moulay Anwar Sounny-Slitine Thanks! I did some more reading and you are correct about the units cancelling each others... However, there a correction coeff. that is different for each band, which would affect the resulting NDVI value if not taken into consideration. I still find your video helpful:) Thanks again!

[@hollisterhurt5629]

Great instinctual video, I can really use this in my occupation as a land manager…Thanks! Is there a way to use this to classify specific tree species?

[@sounny]

If you had more bands of imagery you can do an image classification. But you will need high res, both spatial and spectral. Look up image classification. You also would need some field samples of where known tree are. It is possible but maybe needs a lot of cash to get the right data to do it. Unless you want something very coarse.

[@hollisterhurt5629]

Moulay Anwar Sounny-Slitine It always seems to takes money! If I can ask a simpler question, what is the best way to analyze graphical between Hardwood species to pine species?

[@punithaperiyasamy7199]

thanks, good lesson

[@FarhanAzama]

Just wondering if we can calculate total vegetation using google earth?

[@sounny]

I think you can with earth engine.

[@ajithcsnanjegowda9738]

this question is not related NDVI , but my question is what is the difference between geographic co-ordinate system and projection co-ordinate system

[@6600mamal]

Hi Moulay, May I ask how I can download such satellite images for Los Angeles or Europe? I will appreciate your response.

[@sounny]

+Mohammad Taleghani landsatlook.usgs.gov is a great place to get landsat images for anywhere in the world. There you would use the near infrared and red band, here we did the example with airphoto.

[@Dr.Fitsum_Teshome]

+Mohammad Taleghani you can check Global land cover facility(glcf) or Earth Explorer.

[@gopalbharvad]

i want to calculate a total green area of my state .only green area. is it possible through NVDI ?

[@sounny]

It sounds like what you want to do would be more image classification then NDVI. If you had a mosaic image for the entire state, then you can making a training sample signature file for green and not green and then run a classification for the entire image to find out how much of the image is green not green.

[@maksimfesenko2255]

and sometimes when i upload data it tells me that the pyramid is not sufficient , and asks me if i want to build it . What is that ?

[@sounny]

Pyramids are just an image file that ArcMap makes to improve draw speed on the screen. You should make them. It will not impact your results or your exported maps.

[@rayner3975]

Hello everyone, Do anyone know how calculate the sedimentation index in water bodies using Landsat images and Arcview as GIS tool ?

[@ahcalori2]

Hello, congrats for the excellent tutorial. I already had a NDVI image captured from a modified cam. How can I proceed in this specific case? Thanks

[@sounny]

I am curious which camera can do this, please let me know. If you are just getting red and ir band apply the formula on the georefed image. If it is already applied then you should just georeference the image. I have another tutorial that shows you how to georeference on my channel.

[@ahcalori2]

Hi, thanks for the reply. I wrote wrong. My cam just take pictures in RED and IR bands. It is NIR and not NDVI.

[@sounny]

you should follow the same steps, add the ir and r band then apply the formula with raster calculator. Is you photo shot from above or on the ground. If on the ground don't worry about the georef. If from above then georef.

[@ahcalori2]

Thanks again. Can I work with an image in ".jpeg" format?

[@sounny]

yes should work fine.

[@bakablues]

Thanks for this tutorial. I just got an output with a low value of exactly -1 and high value of 1. What does this mean?

[@sounny]

That would mean that you got the full range, which is pretty rare, but there must be some parts of the image that has full IR and no R and vice versa. Would be interesting to know what were the objects that got those readings.

[@bilguuntserendavaa3850]

my landsatt image just 3 bands ?

[@SitiLiyanaZulkafli]

Hi, I wonder if you/anyone know how to get ndvi using modis (MOD13Q1) image in Arcmap? I really appreciate any help/response

[@adzdimapilis5819]

Hello. What LandSat image you used in making NDVI? thanks :)

[@sounny]

Adz Dimapilis I used an aerial photo. you can take similar approach with Landsat.

[@AngkorEntertainment1994]

Hello guys, i wonder about your generated NDVI value. Why it is not range between 1 to -1 ? Anyone can explain about this ?

[@sounny]

Angkor Entertainment it is very close values to -1 and 1 just this area does not have the extreme values.

[@mohamedhassan-dx7ko]

thank u for that get the picture on the other hand say i have images for every day in the past ten years is a software that can handle processing this images at once or more than dozen at a time

[@yoonasam1]

My NDVI image can't "open attribute table". Please help me How can i solve this...

[@sounny]

since it is a double value, float, it won't have an attribute table. Only integers have that, but you can use the identify tool to see the values of individual pixels.

[@maksimfesenko2255]

Great video ! learn a lot from it ! as im very new to this, i have a question . i have downloaded some imagery from landsatlook.usgs.gov , The imagery i have downloaded are 2 different kind of formats . 1) LC81760222013183LGN00.tar.gz when you choose "Get Landsat Data ( which does not show up at all when i try to add data to my arc map 10.1. and 2) when you choose Get Landsatlook Image you download a zip file with a few file formats but the one that arcgis lets you add is .jpg. So i opened it up and followed your video . Once i got to the band symbology . I realized that while you can choose 4 bands in the video , i only have 3! dont have the fourth one ! Can i still figure out the NDVI with 3 bands ?? also when you download the images you have an option of what sensors to use , could that be a problem that im not choosing the right censors ?? they have like ETM_SLC_OFF, OLI , MSS , and so forth..... Please help ! and maybe you know of another source where i can download imagery for NDVI purpose ? Thank you very much in advance for your help !

[@sounny]

It is important to get the IR and the Red bands. You should get those two, if you only got three bands it might be possible that you have Red, Blue, Green. You need to check that out. Normally the gz file is a compressed folder. You should expand that folder and look at that you have inside it. I think the JPEG might just be a preview image of what is in the GZ. The GZ file might have all 7 bands. So you want to look at the meta data to figure out what the red and ir bands are. I think it might be band 3 and 4 respectively but not sure. Good Luck!

[@maksimfesenko2255]

Yes the 3 bands are red green and blue . So how would i switch them around to get the IR or natural color ? and where can i learn about this ? those combinations with the different bands ?

[@sounny]

Maksim Fesenko Yeah you should look into the gz file, get winrar and unzip it. the jpeg is just a preview. Anyways you should not use the JPEG for this because the image compression loses some of the values you need,

[@InweTasadur]

Moulay Anwar Sounny-Slitine Had the same problem. You downloaded the wrong data. You need a 1 level product, which you have to order (free), it takes a few hours. They are about 200 mb large zip files, which you have to unzipp twice. The seperate bands are in there.

[@sounny]

Thanks for the clarification! In this video I was using Aerial Imagery. NASA images work a little different but the NDVI calculation is the same.

[@imranahmedkhan8130]

I am not getting voice.