Satellite Imagery Clouds Cover Removal in ArcGIS Desktop

[@GEO2004]

better, easier, more accurate, more precise, and full utilization of QA Band Mask for Satellite Imagery Cloud Removal and Cloud Correction, Check this Newest tutorial : kzbin.info/www/bejne/aoGvpqaQnr-Vfbc

[@suttonkellan3475]

i know I am quite randomly asking but do anyone know a good place to watch newly released movies online?

[@reneharrison1581]

@Sutton Kellan i use FlixZone. You can find it by googling =)

[@alvinporter1283]

@Rene Harrison Yea, been using flixzone for since april myself :D

[@suttonkellan3475]

@Rene Harrison thanks, signed up and it seems to work :D I appreciate it !!

[@reneharrison1581]

@Sutton Kellan glad I could help :)

[@eddisonjose]

Hello thank you very much. Once the max values ​​are tiled, I assume the radiometry is affected. In that case is it no longer possible to use a supervised classification for land uses?

[@GEO2004]

Depends on the classification purpose and schema, if you just want to classify vegetation, water, bare soil and builtup area, I think it is still possible

[@eddisonjose]

@@GEO2004 Ah ok thanks, but to generate spectral signatures is not it possible? Thank you

[@GEO2004]

I guess it is still possible. It is only you needs to perform atmospheric correction and get the surface reflectance data prior to clouds removal. Atmospheric correction theoretically should also remove seasonal illumination variation between acquisitions

[@afeworkalemayehu9434]

Amazingly, I heard that cloud can't be removed from images. This is wonderful presentation for me. Thanks!!

[@GEO2004]

Glad it could help Yes, it is can, but the imagery result can't be treated as single date imagery anymore. There are some consequences like inconsistent landcover due to the possibility of landcover changes between dates, or any other effects that arise due to different imagery merged into single dataset

[@jaymakorani4019]

After applying this technique, is it possible to perform grey levels-based thresholding with resultant image?

[@GEO2004]

Sure

[@bb-ks6ow]

satelit imagery guna band apa?

[@GEO2004]

Macam macam

[@dirham-ugm]

Hi there thanks for the help! But here I got an issue on the last step, when I tried to mosaic the two dataset altogether it tends to display rather blueish than the original satellite imagery (without any image correction). I've tried to reduce the brightness on the third band (blue) of my mosaic data, but it ruined the whole picture. Is there any function that we could use to adjust only the masked part on the final mosaic raster?

[@GEO2004]

Agisoft has a feature to create mask layer, however I don't know if it can be used for local brightness adjustment, I guess you can try it out

[@muhammadtaufikiqbal]

useful for landsat band 1-7??

[@GEO2004]

Certainly

[@ozdebakak4140]

Hello. I'm wondering whether to use it to determine changes in Sea Surface Temperature in the case of pre-post seismic events. After seismic events, generally, a temperature increase is expected, of course sometimes not to this. I don't want to do wrong analysis, thus data interpretation. Do this image processing acceptable? Thank you for your video.

[@GEO2004]

I am not really sure, I guess you must verify it by yourself

[@ozdebakak4140]

@@GEO2004Thank you very much for the quick response to my question and your video. I'll try to do this :)

[@meantalk8134]

Hi! What should I do if my imagery has five bands?

[@GEO2004]

Doesnt matter as long as the imagery is loaded into arcgis using original metadata (which means the particular imagery metadata format is supported by arcgis)

[@mrs_park8803]

Can i follow this step for landsat 8?

[@GEO2004]

Yes you can

[@raufmammadli2027]

Hello, where did you get the number 1200? How can I get this number?

[@GEO2004]

by inspecting the imagery, you can use identify tool and repeat clicking in the zone between clouds cover and actual landcover, on that area you will find the good value that separates clouds and non-clouds for every bands in your imagery

[@raufmammadli2027]

@@GEO2004 I have 7 bands showing. but you have 3 bands. how should i do Help me. How can I contact you?

[@jesfermamon4894]

is this good enough for calculating ndvi?

[@GEO2004]

Depends on analysis purpose, I already explain it in other comments

[@muhammadtaufikiqbal]

how to know the maximum value for each band??

[@GEO2004]

Depending on the sensor type, some have 8 bit which the maximum is 255, some has 11 bit or even 16 bit that obviously the max number will be much larger

[@ahmadzulfadli3115]

Hi sir, I would like to ask some question from you sir. Both image that you used have different date, so how long their date gap? Is it this step is suitable for rural area only? because my satellite image Sentinel 2B is urban area and if the date gap is too long, they will have some changes of the features if I want to do LULC classification. Am I right sir?

[@GEO2004]

Yes you are right, the gap should be as narrow as possible

[@sharmisthabhowmik9599]

Can it be done with single bands one by one? Like band 1 then band 2 n do on

[@GEO2004]

yes it can, if you prefer doing it band by band, there are more arithmatic operatior you can use, but it is more tedious for sure

[@MuanteMT]

Hi! Im fairly new to this area of study. Nonetheless., I was wondering whether this type of cloud removed data can be used to determine NDVI/NDWI accurately?

[@GEO2004]

it will depends on your operational definition of "accuracy". I will give you an example. Lets say, there is an analyst who want to calculate forest cover loss using NDVI values on tropical area for 5 years range. For this kind analysis, the analyst made annual NDVI dataset coming from the aggregation/compositing of imagery taken for all the date along the particular year. Because tropical area usually are not free of clouds cover, Clouds removal would be important for the analyst workflow. Because that is the only way to get a clear imagery as the forest cover classification requirement for A YEAR. the processed cloud-free would resulting in very good and accurate forest classification representing one year condition. On the other hand, there is other analyst who want to use NDVI to monitor agricultural crop cycle, measuring crop yields, or any other precision agriculture application. Using NDVI coming from cloud removal that involves imagery from few months or weeks will give them misleading analysis result. Why we called it misleading? that is because crops usually have shorter temporal dynamic of phenological cycle, so if the analyst force to use cloud removal, their result would be far from accurate, it won't capture the precision needs from time series imagery analysis.

[@MuanteMT]

Suppose i want to analyse Agricultural drought over a certain monsoon season. So in that case, is it practicable if i use cloud corrected images?

[@GEO2004]

Yes, but pay attention to the time frame, if the images acquisition date are too wide, you will lose the vegetation dynamic you are looking for

[@MuanteMT]

Okay. Thanks a lot👍🏻

[@bb-ks6ow]

mcm mana nak tau range cloud cover pixel value?

[@GEO2004]

Inspect in the imagery between cloud cover area and no cloud area

[@belindaaulia3437]

Mau tanya, di dalam mask function itu band sy ada 7 bukan 3 seperti di contoh tutorial, jadi bagaimana mengisi nilai maximum cloud pixel valuenya ya? terima kaish.

[@GEO2004]

Tinggal diinspeksi di setiap band, tapi nilai maximum cloud rata2 sama di tiap band, kecuali band termal

[@dimsrasau2453]

Untuk Nilai 1200 dan 800 itu di ambil dari yg bebas awan atau yang berawan ya mas?

[@GEO2004]

Nilai batas minimal maksimal yang tidak berawan

[@annanda7319]

@@GEO2004 halo, ingin bertanya. nilai batas minimal dan maksimalnya didapatkan dari mana ya?

[@GEO2004]

Cek manual di seluruh liputan citra, kalau hasil kurang pas, bisa diulang sampai ketemu batas nilai paling mewakili dan paling presisi

[@floratech9127]

Can u help me how to download satellite images for cloud removal

[@GEO2004]

There are numerous video in KZbin which demonstrate how to do it. My advice is just make sure there are no clouds in same area between imagery or else you won't get cloudless image.

[@hasanudin898]

halo saya mau bertanya, setelah apply mask setting layer yang keluar hanya awannya dan selebihnya abu2 bukan seperti di video yang masih terlihat citranya. Gimana ya untuk solve the problem? thanks

[@GEO2004]

apakah anda menggunakan citra dengan format seperti yang ada di video?

[@faridanurrahmah4203]

halo bang, saya baru belajar arcgis, mau tanya cara buat data image. tif nya gimana ya?

[@GEO2004]

Kalau baru belajar, jgn langsung praktekin ini, pelajari yg lebih basic dulu

[@faridanurrahmah4203]

@@GEO2004 permintaan dosen bang, buat peta tapi awannya dihilangin

[@arnolddaniel6529]

Gimana kalo band nya ada 11, harus semuanya 1200?

[@GEO2004]

Ya tidak, emang citra apa yang band nya 11 ? Kalau sentinel 2 atau landsat 9 kan tidak semua band dipakai.

[@87_adityanurfauzan83]

Permisi min, saya mau membuat ndvi dalam waktu satu bulan. Namun dalam jangka waktu tersebut banyak sekali awan yang menutupi jadi apakah awan tersebut bisa di hapus atau tidak ya ? Karena seingat saya jika tutupan awan cukup banyak akan mengurangi tingkat keakuratan citra satelit jika bisa setelah pemotongan awan apakah ada step lagi ? Mohon bantuannya min terima kasih

[@GEO2004]

Tergantung jenis analisis yang akan digunakan dari ndvinya

[@techforkannada7533]

its greats