Відео

Belum bisa setau saya

@@geografif kalau dari tools dashboard yg lain kira-kira ada yg bisa nampilin kaya gitu gak min, misal tableu atau power bi?

@@solahudinalayubi4597 setau saya kalau berbasis GUI fiturnya terbatas. Kalau mau berbasis script bisa coba Observable.

Yes, separate file for each target class as shown at 0:52

Halo, apakah format data yang diinput sudah sesuai seperti 3:48?

Hi, those data/layers are aerial images that came from a project, so they are not open. But you can always check OpenAerialMaps for publicly available dataset. Example tutorial: ua-cam.com/video/BCTcscArHbE/v-deo.htmlfeature=shared

Saya belum coba, tapi ada yang pernah membahas di forumnya: community.opendronemap.org/t/process-for-a-gpu-node/11743/8 Kalau setup gpu-nya sudah oke, bisa run webodm via ./webodm.sh start --gpu

By dragging them into qgis from my local storage. Those are aerial image layers

Hi, if you have an aerial video, you can split the video frames into static images. That way, you can stitch them into continuous image mosaic and create a Digital Surface Model (DSM). The DSM can be used to create a 3D model. The challenge would be in giving the georeference to the stitched image mosaic as video recording does not have that metadata.

Hi, thanks. Glad it worked for you.

Halo, machine learning/deep learning bertujuan untuk 'belajar' atau mencari rumus untuk memprediksi target (y') berdasarkan data input (x) dan target (y). Dalam kasus segmentasi semantik atau klasifikasi per piksel, x adalah citra dan y adalah label (dengan dimensi/ukuran yang sama dengan citra). Pada video, citranya adalah 2 band polarisasi Sentinel-1. Sedangkan label adalah peta tutupan lahan sawit dan non-sawit. Yang anda jabarkan adalah input citra bukan label. Walaupun memungkinkan untuk menambah indeks NDVI ke dalam stack raster RGB-NIR, sebenarnya tugas deep learning (dan yang membedakannya dari machine learning tradisional) adalah mencari representasi fitur otomatis (dari band spektral asli). Dengan demikian, feature engineering seperti menyusun indeks NDVI tidak diperlukan lagi. Citra / data input itu nantinya akan diproses menjadi (kelas target) apa? Itulah yang seharusnya menjadi labelnya. Label bisa diperoleh dari digitasi manual, semi-otomatis (thresholding indeks dan segmentasi), maupun data/peta yang sudah ada (tutupan lahan, building footprint dsb.). Nantinya dapat dirasterkan dan dimensinya dibuat sama dengan pasangan citranya. Semoga bermanfaat.

@@geografifterima kasih banyak ka, penjelasannya sangat detail

Dicoba saja langkah di video dengan mengganti nama layer sesuai interest

@@geografif hanya shp deforestasi 2023 saja yang bisa di download kak

​@@derifirmanto7093setelah menit ketiga kan ada cara unduh layer lainnya 😊

Which collection did you try to load?

See it yourself. The link is on the description 😀

@@geografif It's unavailable

@@santoryuu6807 forgot to make it public. But if you set the Looker report to be public, then yes, the images can be seen by anyone since the image's URLs are also public. See the spreadsheet link in the description.

Meant to add subtitle but ive been busy. Thanks for the suggestion.

You can either try the older version (e.g. 8) or do it by scripting: ua-cam.com/video/_V0X1LdSta4/v-deo.htmlsi=o6LnCzwZ3YhMuuee

@@geografif i have tried the older version.. i found sen2coral, but i couldn't add the sentinel product, always show that "unable to read metadata from MTD bla...bla...

Try to open the individual .jp2 file if you downloaded the data from Copernicus hub

Url tambahan supaya webpage ngeload json yang bisa diunduh sebagai layer SIG

Masih belom berhasil bang, saya ikuti step2nya. Apakah masih berlaku ini stepnya

@@boylatanza sepertinya aksesnya sudah ditutup

Ia bang.. semoga ada updatenya lagi nih cara terbaru bang. 🙏🙏

Saya tidak tahu

Halo, data terakhir tahun 2021 saat rest servicenya masih dibuka

Go to Extension > AppScript

It comes with the spreadsheet. Check the link in the description.

You have to be clear what didnt work for people to be able to help you

Hi, they are already in the video :)

Boleh kirim link ke scriptnya

No, that needs a different procedure

Which version are you using?

@@geografif oh no, it is not compatible with SNAP 10. I want to testing mapping seagrass with Sentinel2. do u know how

I've never tried it on newer version of SNAP. The video was using SNAP 6 or 7. Anyway, you can always download the older version on their website I believe

@@geografif lets me try..thz

Did it work?

Hi, did you make sure each netcdf file has the same extent?

@@geografif I did not I assuemd they are. How would you Check this and sollen the issue?

@@user-zs8wl1ut2t you can check this resource: opensourceoptions.com/netcdf-with-python-netcdf4-metadata-dimensions-and-variables/ to explore your netcdf data first before doing further step e.g. merging.

Halo, silakan email ke akaefgefan@gmail.com

Im not sure I understand your problem. The idea of using URLs is to keep your images somewhere else so you dont have to put it manually to your sheet (or whatever database table you are using as most of them dont support image file type in their column). I dont see the point of using the method I show for only 'an' image.

I believe so. Obtaining those parameters requires a little bit modification to the 'surface_reflectance' function so it dooesnt just provide image as an output but also the value of the two parameters.

@@geografif Do you have the tutorial to derive (Pa) and (ωo) values? I'm currently working on my Thesis project, which requires (Pa) and (ωo) values. It would be a major help if you have.

Glad it can be helpful. Please subscribe to support the channel, thanks!

Jonah, Gideon, Rebecca, Delilah, Ric

Yes it can. Ive been planning to do that for a while now but havent really got time

@@geografif its me again... is there any chance we can chat by email or something else to solve some doubts? also if youre available to mentorship me in DII on GEE... i can pay. Also i want to know if Water column correction is avoidable?? or it is a mandatory step for habitat clasification?

@FranciscoJavierChaconSantizo hi, water column correction is not mandatory. Sure, we can arrange mentorship. You can contact me at akaefgefan@gmail.com.

Make your AOI smaller

Which snippet gave you the error?

@@geografif Hi, l solved problem. It was path error

@@bextla3737 nice 😀

¿En qué minuto?

Soalnya saya nyoba nyari beda dikit / kurang nulisnya datanya gak muncul samsek,🙏

menit mana ya yang dimaksud?

Hi, the notebook is included in the github repo (url) shown at 1:12

'maxPixels' is part of the codes that tell the system to export an image to Google Drive. When you click the 'run' button, it runs the whole codes, hence the processing indicator in the Tasks bar. The format can be specified in the code together with maxPixels or by clicking the button of each tasks. Once exported to your Google Drive, you can download the image to your computer for offline uses.

@@geografif thank you, very easy with your video...

Plugin dari BuntingLabs ya? Nanti deh dicoba dulu

I don't manage the place

Tergantung kecepatan hardware prosesor. Sebelumnya apakah sudah pernah mencoba dengan sampel kecil dan berhasil?

@@geografif sebelumnya, belum pernah mencoba. Saya langsung mencoba dengan foto tersebut

Normalnya tidak sampai sehari. Ada baiknya mencoba terlebih memproses dulu sampel kecil agar tau apakah instalasi sudah benar. Beberapa sampel bisa ditemukan di kolom komentar video berikut: ua-cam.com/video/aXg4udoZ5-4/v-deo.htmlsi=qPWQ1p1IbKc70Wkw

Ya wms. Beberapa data tersedia di portal lain seperti ini: ua-cam.com/video/wXTjk2BmnG8/v-deo.html

@@geografif raster, dibatasi juga skala 1:100.000 baru dirender. capek saya, maksudnya apa sih sharing data publik dengan cara aneh. Itu web SIGAP sudah setahun lebih punya halaman login untuk analisis spasial tapi tidak bisa masuk, registrasinya malah error terus. Tiap tahun pembaharuan data hutan tapi tidak dipublikasikan dan digunakan dengan baik, orang ga boleh sembarangan masuk kawasan hutan. Tapi mau tau batasnya dimana dibuat repot. Giliran udah masuk pinalti sana-sini. Penasaran saya tender sama siapa itu web SIGAP.

Haha. Protesnya ke akun Kalehaka atuh pak :)

Bahasanya tetap javascript ya. Matplotlib di GEE itu adalah library di javascript untuk memanggil palet warna Viridis. Kalau tidak mau menggunakan palet warna preset seperti Viridis, tidak wajib pake Matplotlib. Pada bagian visualisasi bisa diset misal dari putih ke hijau untuk nilai rendah ke tinggi.

Email saja ya di akaefgefan@gmail.com

@@geografif oke kak Terima kasih

Please follow along the video, thanks

Bisa