Object-based Image classification in QGIS || OBIA !! || A complete Tutorial
Object-based image analysis (OBIA) involves pixels first being grouped into objects based on either spectral similarity or an external variable such as ownership, soil or geological unit.An Object Based Image Analysis (OBIA) approach to image classification segments an image into homogenous spectral regions.
This differs from standard image analysis methods that classify each pixel in isolation; the OBIA approach generally produces a cleaner map product, more meaningful to human users.
If you'd like to support our journey and help us provide even better GIS and Remote Sensing insights, consider contributing through 'Buy Me a Coffee'. Your support ensures we can continue creating tutorials, sharing tricks, and making our content as informative and enjoyable as possible.
🙏 Contribute💰💰 Here: www.buymeacoffee.com/jessebuy...
🙏 Contribute💰💰Here: www.buymeacoffee.com/jessebuy...
Пікірлер: 194
If you'd like to support our journey and help us provide even better GIS and Remote Sensing insights, consider contributing through 'Buy Me a Coffee'. Your support ensures we can continue creating tutorials, sharing tricks, and making our content as informative and enjoyable as possible. 🙏 Contribute💰💰 Here: www.buymeacoffee.com/jessebuyungo 🙏 Contribute💰💰Here: www.buymeacoffee.com/jessebuyungo
Te amo mi rey. Sos un Genio!!!! Mis aplausos para Usted!!!!
Super helpful and clear instructions, can't thank you enough for making this.
Very good instructions and it really helped me. Thank you!
That is a lot of work! Thanks for sharing
This is a very good tutorial, thanks for posting.
This have helped me a lot in my work. Thanks a lot!
Great Explanation! Thanks!
Thanks for the help. Signed by the Water Sciences Master's degree from Lyon (France).
Thank you very much, clear, new and good tutorials: I have used a 0.15*0.15 high resolution digital orthophotos following your step, I get a good result. Thank you very much again
Great. Can you make a separate video for accuracy assessment as it is necessary after classification to prove our results?
Super 🥰 ! Thanks... 🤗
Hello, I have some problem with TrainVectorClassifier. When I choose "joined layer" as Input Vector Data I see a messege: "Unable to load selected layer or table. Dependent field cannot be filled in". How can I solve that? Thanks for your help.
Professor, how can I create a DEM elevation model file from the surface of the external appearance using the Qgis Orfeo Toolbox plugin? Please answer me, and thank you very much, Doctor.
Thank you so much, sir, it was really helpful!
@jessebuyungo
2 ай бұрын
Glad it helped!
Hello Jesse, I face a challenge, during segmentation the message "Self-intersection at or near point" appears, and I can not continue the segmentation process. how to fix this problem?
👍👍👍 Am closely following once I try it out will have my reactions
@jessebuyungo
2 жыл бұрын
No worries Isaac. Make sure you have a working installation of the OTB tool box
Hi Jesse, when trying to run the TrainVectorClassifier I get the Error message that directory does not exist... What would be the work around for this?
What's the sourse of your satellite image. And how did you get your resolution to be that good?
Dear Sir, when I run OBT_Vector Classifier, it does not produce any files. I think it is a problem reading the *.model file. How can I check this? Please can you help me? thank you very much
@natcain6549
Жыл бұрын
Hi, I am having the same problem
@giacomoquattrini8254
Жыл бұрын
I am having the same problem
@salvadorrenelealramos8918
Жыл бұрын
Hi, I am having the same problem
@martingrosso8689
Жыл бұрын
I am having the same problem
@pietropianca
10 ай бұрын
I have the same problem. How can i solve it?
Is it possible to have another tutorial on how to use the Segment Anything Model in QGIS?
hey Jesse thanks for the video! I have a question in Train Vector Classifier part, if i have mean_3 and stdev_3 do i also put that? also why do we use mean and stdev in Train Vector Classifier?
@jessebuyungo
7 ай бұрын
Steve, Mean are parameters that are computed during the compute statistics stage, they are related to the mean pixel and the std of the pixel in each segment.......that's y the number (n) depends on the number of bands the image has.
Hello Jesse, thank you for your great tutorial video. I have a question regarding the OTB plugin. I couldn't find the plugin in QGIS 3.16.
@jessebuyungo
Жыл бұрын
Please you need to install otb, watch my tutorial on how to
The OBIA_model.model doesn't want to save on our computer as an output-file during the TrainVectorClassifier-step. Do you know what the problem is?
Great video.
@jessebuyungo
2 жыл бұрын
Thanks
Does anyone know how to increase the number of iterations for the SVM classification method using the train vector classifier
Hello, Jesse. Thanks for the great video. I encounter the problem, that my segmentation results generate dozens (more than 1000) self-intersecting polygons. With those I was not able to proceed the process, so I tried to just use the "fix geometry"-tool an so far it worked out fine.
@jorritxl
2 жыл бұрын
Hey hannes, i had the same error, were can i find the 'fix geometry' thing?
@hannespillsticker5852
2 жыл бұрын
@@jorritxl Hey Jorrit, you can find the "fix geometry"-tool in the qgis toolbox. For further explanations, check out the documentation or here: www.qgistutorials.com/en/docs/3/handling_invalid_geometries.html. Greetings
I have a question there are three mean and statistics calculations Does that signify 3 bands of zonal statistics, Which means otb zonal statistics can process multiband zonal statistics calculation is that correct ??
i have 6 class_id´s but predicted only classifies everything as the same (i.e. 1 of 6), even those polygons which have been given class ids. potential issues might be that im using sentinel-2 data, which results in 9x mean and stdev fields. additionally, after the join by location, are you using a filtered dataset for the training (ie only those polygons that have been assigned class_ids) or does your training dataset contain all polygons with both class id´s assigned and polygons with class id NA?
@olliebartlett5029
2 жыл бұрын
2021-12-01 15:29:08 (INFO) VectorClassifier: Default RAM limit for OTB is 256 MB 2021-12-01 15:29:08 (INFO) VectorClassifier: GDAL maximum cache size is 813 MB 2021-12-01 15:29:08 (INFO) VectorClassifier: OTB will use at most 8 threads 2021-12-01 15:29:08 (INFO) VectorClassifier: Model loaded 2021-12-01 15:29:08 (INFO) VectorClassifier: mean used: [0] 2021-12-01 15:29:08 (INFO) VectorClassifier: standard deviation used: [1] 2021-12-01 15:29:08 (INFO) VectorClassifier: Loading model log files from vectorClassifier output.
@jessebuyungo
2 жыл бұрын
May be you Didn't configure the training very well. Could you please share a screenshot of the error?
@asrafs.hadjiali2454
Жыл бұрын
@@jessebuyungo greeting of peace, I have same problem.
Hi, Thank you for your great video. I have to face a problem with the "Train Vector Classifier". I received an error message " TrainVectorClassifier: itk::ERROR: FieldParameter(0000022C97C15A30): Single selection mode is on, but there are 2 selected items ". Please help me.
Hi great tutorial right there. If you don’t mind answering, i would love to ask. Can the analysis be done at urban/regional scale spatial unit? Because from your tutorial, the analysis is done at neighborhood scale
@jessebuyungo
2 жыл бұрын
Yes please you can do the analysis at a regional scale
@hernandbagaskara
2 жыл бұрын
Jesse Buyungo does it require survey or sampling ?
@jessebuyungo
2 жыл бұрын
@@hernandbagaskara sampling is fine
How to fix this error using trainvectorclassifier ??? (FATAL) TrainVectorClassifier: itk::ERROR: FieldParameter(0000025D49A515F0): Value ['_mean', '_median', '_stdev'] not found in the list of choices: DN, _count, _sum, _mean, _median, _stdev, _min, _max, _range, _minority, _majority, _variety, _variance, id, class_id.
I cannot find OTB Zonal statistics tool on my QGIS 3.22
Great job @Jesse!! I would like to look at the Accuracy assessment results. How accurately did the algorithm predict the class?
@jessebuyungo
2 жыл бұрын
Well I you like to do accuracy assessment, make sure you pick testing samples and the same way you did for the training data. And supply it to the algorithm during the training. It will give you a confusion matrix which you can analyze ......and generate producer, user, overall accuracy parameters, etc
which version of qgis did u use
Hello Jesse, thanks for creating such a valuable tutorial on OBIC. It's been incredibly helpful. I was wondering if you could kindly share the raster image you used.
@jessebuyungo
2 ай бұрын
Wow, this is a Pléiades satellite image taken in Kampala CBD, Uganda
@mominfpm1
2 ай бұрын
@@jessebuyungo Thanks. I could not find it on EarthExplorer. I wonder if you still have that image.
@jessebuyungo
Ай бұрын
It's a commercial satellite image. So you have to buy it at a cost
i encountered an error whilist activating the otb it said there is a typeError in the python so how can i solve that
What means "More than 1000 errors or warnings have been reported. No more will be reported from now." ? Please help me.
Thanks for Video Jesse. Everything works up until the VectorClassifier part of the video. When I run VectorClassifier I get the following: 2023-03-31 21:22:23 (INFO) VectorClassifier: Default RAM limit for OTB is 256 MB 2023-03-31 21:22:23 (INFO) VectorClassifier: GDAL maximum cache size is 398 MB 2023-03-31 21:22:23 (INFO) VectorClassifier: OTB will use at most 8 threads Could not read file C:\GIS\back Any idea what the issue is?
@martingrosso8689
Жыл бұрын
Idem problem, no .model no .txt 2023-06-25 20:00:53 (INFO) VectorClassifier: Default RAM limit for OTB is 256 MB 2023-06-25 20:00:53 (INFO) VectorClassifier: GDAL maximum cache size is 788 MB 2023-06-25 20:00:53 (INFO) VectorClassifier: OTB will use at most 16 threads Could not read file C:\Users\Usuario\Desktop\Nueva
sir is there any way you could demonstrate sun glint and water column correction and OBIA using planet data in qgis or arcgis for benthic habitat mapping?
@elviraarmenio9149
Жыл бұрын
Hi, I have the same problem. I am trying to apply classification to images acquired with UAV on the water surface (full motio video) to detect posidonia. Have you had any feedback to your question? can you give me any suggestions please? thanks
Dear Sir, thank you so much for this nice video! One question: how do you do accuracy assessment after the classification? Is there some kind of confusion matrix that I can get in QGIS after the classification? Thanks for your help
@jessebuyungo
2 жыл бұрын
Yes please, prepare test and train data shapefile differently..... during model training...input the test data tooo.....and a confusion matrix will be output
@majetetube2023
2 жыл бұрын
Dear Karina, I also Face a challenge to perform an accuracy assesment in a train vector calssifier. Have you got a solution to accuracy assessment? please share me!!!
@user-be4lf1ng2f
10 ай бұрын
@@jessebuyungo Hello Sir. Can you make a separate video for accuracy assessment as it is necessary after classification to prove our results?
Is it possible to reuse the segmentation on other similar satellite images? So you dont have to set new datapoints on a new image?
@jessebuyungo
5 ай бұрын
yes please, you have to segment the imagery each time
I have this series QGİS 3.20.3. Is it possible with it?The segmentation tool is not found when searching in the toolbox.Please help me
@jessebuyungo
2 жыл бұрын
Please have you installef the OTB plugin already??
When I do categorized after the classified output, i get only 1 class; though i have created 5 classes during the training data/sample. What seems to be the problem?
@jessebuyungo
2 жыл бұрын
Please open the attribute table of that layer and check first the predicted classes, if you see all the five classes then go on to the symbology and categorize, Alternatively if you see only one class predicted then please you need to recustomize your training parameters and re-run the algorithm.
@olliebartlett5029
2 жыл бұрын
did you resolve the problem? i haver the same issue (only one predicted classs, even for polygons that have been classified as a different class)
I still get the "Could not load selected layer/table. Dependent field could not be populated" error even when I saved joined_layer as shp file, can you please help?
@natbrzezina
9 ай бұрын
I have the same problem, plus my joined layer has only 1 set of mean and stdev fileds
Thank You, Which Ver. of QGIS is required ?, How do I check that OTB Tool box is installed ?
@jessebuyungo
2 жыл бұрын
You're welcome, For this specific tutorial, I used Qgis Hannover, after installing the OTB, you will begin seeing it's algorithms in the tool box menu
@davidbehr7865
2 жыл бұрын
@@jessebuyungo Hi, I have installed OTB (7.3.0) and QGIS (3.14), Win-10,64Bit. However, the Segmentation menu looks differant than you presented (LargeScaleMeanShift) , is it only a change in OTB UI ?
@jessebuyungo
2 жыл бұрын
@@davidbehr7865 Ohh could you please share the snapshot in an email please @jessebuyungo48@gmail.com
Thank you for this video, can we use google street view or image, and use Qgis for image classification without using deep learning?
@jessebuyungo
2 жыл бұрын
No please. We have to use Aerial or satellite imagery
@simonhanssens8459
8 ай бұрын
why isn't it possible to use an image you took by yourself? @@jessebuyungo
Hi, can you help me, I couldn’t find the vector classifier output in the folder even though the process is finished
@pietropianca
10 ай бұрын
Hi, i have the same problem. Did you solve it?
if I have sentinel 2 data, which data should I enter? are all bands or bands stacked? sorry if my english is not polite sir, i use google translate. Thank you sir
@jessebuyungo
2 жыл бұрын
No worries,. Its fine. Sentinel 2 stack B,2,3,4 to make RGB color image
Thank you sir for the tutorial. I want to ask you how to extract lineament using obia?
@jessebuyungo
Жыл бұрын
Well, the procedure is the same
Thanks for your video. I have a problem with the last step ''Vector Classifier''. When I run the process, it doesn't make me the classified_output.shp. Do you know what happen? I need to do a special step?
@Technologic911
11 ай бұрын
I had to the same issue. Solved it by making sure that the file path didn't have any spaces in the name. Had 1 and replaced it with an '_'; things work!
@user-nd1rs1po6g
11 ай бұрын
@@Technologic911 Thanks, it isn't the problem. It was the qgis version because the otb version needs a upgrade qgis and just I download the new qgis
@aislingoflynn6167
11 ай бұрын
This is happening to me, what version of qgis and otb did you need to get it to work? Thanks
@user-nd1rs1po6g
11 ай бұрын
@@aislingoflynn6167 3.20 or upper
@pietropianca
10 ай бұрын
Hi@@user-nd1rs1po6g i have the same problem. When i run 'vector classifier' it doesn't create the classified_output.shp file. I downloaded QGIS version 3.32.1 Lima and OTB version 8.1.2, but the problem is not resolved. What can i do?
Thanks for this great video! I am trying your method and so far I have no problem from the start until the part of "vector classifier". It keeps showing this: "2021-08-10 22:20:06 (FATAL) VectorClassifier: itk::ERROR: Failed to open GDALDataset file C:\Users\" and it does not produce the classified output layer. I am using QGIS 3.18 Zurich and OTB (7.3) Windows 64bits with OTB master plugin. I wonder what could be the problem? I would appreciate your help
@jessebuyungo
2 жыл бұрын
well could you please share a snapshot of your Input parameters plus the error screen snapshot please
@sarahjane4484
2 жыл бұрын
@@jessebuyungo would you mind if I share it in your email instead here? thanks
@jessebuyungo
2 жыл бұрын
@@sarahjane4484 kindly contact me using: jessebuyungo48@gmail.com
@adeerkhan1
2 жыл бұрын
Whatever file you save, and the folders in which the files are located make sure their names don't have a space, rather use _ instead of spaces.
@sudiptaroy8563
2 жыл бұрын
@@jessebuyungo I have sent you an email regarding the same problem.
train vector classifier 'Could not load selected layer/table. Dependent field could not be populated'. So no sence to watch AFTER THAT AS WELL
Nice video and understandable instructions. please sir, can you share my your presentation slide on this video or pdf.
Great video. Can you share the tutorial data, please? Can I use Google Earth Imagery for this?
@jessebuyungo
2 жыл бұрын
Thank you so much Michael. Well when it comes to the data, you need to download satellite imagery. You could consider downloading freely available sentinel 2 or landsat 8.
@einsor7260
2 жыл бұрын
@@jessebuyungo if I have sentinel 2 data, which data should I enter? are all bands or bands stacked? sorry if my english is not polite sir, i use google translate. Thank you sir
@jessebuyungo
2 жыл бұрын
@@einsor7260 stack b2,3,4
Hi sir I would like to ask if you have a tutorial on how to segment the tree crown using OTB or anything that will segment tree crown using the UAS data
@jessebuyungo
2 жыл бұрын
Try using this approach and check the results please, however this requires high resolution imagery
@princessdorindingal316
2 жыл бұрын
@@jessebuyungo I used this approach but it was oversegmented and there was an error "Self-intersection at or near point". And I also used UAV image so it will have a high resolution. Can you recommend how to segment only the tree crown and the parameters to used for this? please
Please help! Vector classification does not have an output for me, maybe its because of the model? Im using 3.32.1 - Lima version
@jessebuyungo
8 ай бұрын
Please share a screenshot error
@jessebuyungo
8 ай бұрын
Please share a screenshot error
.model file not creating plz tell me what to do
can i use this methodology for a binary classification?
@jessebuyungo
2 жыл бұрын
Not so sure, it may not work, still you can do a multi class classification and visualize it as a binary one...... Quite simple to do
Hey, great work. However, I can´t use TrainVectorClassifier, it keeps showing the same error, "Could not load selected layer/table. Dependent field could not be populated", after loading the joined layer. How can I fix it? Thank y
@jessebuyungo
2 жыл бұрын
Save as .shp
@bidiptaghosal7411
2 жыл бұрын
@@jessebuyungo I have saved it as a .shp file but its still showing the same error when I am trying to carry out the TrainVectorClassifier. How can I resolve this issue?
@jessebuyungo
2 жыл бұрын
Which Qgis version are you using?
@bidiptaghosal7411
2 жыл бұрын
@@jessebuyungo QGIS DESKTOP 3.16.1
@jessebuyungo
2 жыл бұрын
@@bidiptaghosal7411 I don't know now because I use the same version
if I use polygons instead of points for the training sample it the classification goin to be more accurate?
@jessebuyungo
2 ай бұрын
We. Are not using points actually, we are selecting points which we later use to select our polygons
@lorenzoconti7372
2 ай бұрын
@@jessebuyungo got it!! in your opinion what model is the most accurate? Im trying em all but it seems that linear and polinomial are the better ones
@jessebuyungo
Ай бұрын
...I can't say that there is an accurate model, different model will be accurate depending on the given situation, maybe the landscape, the samples, the satellite imagery etc.....
@lorenzoconti7372
Ай бұрын
@@jessebuyungo I undestand, another thing, im trying to do the segmentation with a landsat but it doesnt work, the output is just a big polygon that cover the study area. Do I need to change some settings for landsat? I tried to resample at10 meters but It doesnt work
@jessebuyungo
Ай бұрын
Can I give you a tutorial for that at a discounted price?
can you explain why u use autocad and not shp in the last, i didnt see my file output any help pls ?
@jessebuyungo
3 ай бұрын
Hi there, am sorry, but I haven't used AutoCAD in this tutorial, Incase you are facing challenges, I can take you through under a private arrangement please
@amirouchelouaidi3250
3 ай бұрын
@@jessebuyungo I once again tried all the steps at the final classification stage to open the output shp file, I cannot find it despite the message specifying that the document has been classified, can you review this on your side maybe there was an error
@jessebuyungo
3 ай бұрын
Ohhh sorry chief. Try checking in the directory you saved it
Hii there! the segmentation process shows completed within seconds and there are no results...please help
@jessebuyungo
2 жыл бұрын
Could you please share a screenshot of the error/ result screen
hello, I face some struggle during the "VectorClassifier". The shp is missing at the end. I can find it nowhere and even I try to run it on temporary file, it doesn't show up either. Thanks in advance
@giovanniandrianjafimamy5129
3 ай бұрын
@jessybuyungo I need your help
@jessebuyungo
3 ай бұрын
Hello chief, what does it say after classification?, does it give an error?
@giovanniandrianjafimamy5129
3 ай бұрын
@@jessebuyungo hello, so basically the algorithm work perfectly but I can't find the output. Plus there's no error message after running the "vectorclassifier". I just don't understand why it's missing. I choose the file for the results and everything but it's still missing
@jessebuyungo
3 ай бұрын
I can't understand the problem, maybe we can arrange an online meet and you take me through and I can guide you where you are going wrong
@giovanniandrianjafimamy5129
3 ай бұрын
@@jessebuyungo ohh really, thanks in advance.
Hey great vid, I tried to copy but it resulted in this error: Warning 1: Self-intersection at or near point 145032.5 408201.5, do you know what went wrong?
@jessebuyungo
2 жыл бұрын
Did it stop there or it proceeded???
@majetetube2023
2 жыл бұрын
Dear jorrit, I also face this challenge? have you get a solution?
when the vector classifier process there is an error : failed to open GDALDataset file D, how to handle it
@jessebuyungo
2 жыл бұрын
Could you please share a snapshot of the error screen and we see how to help you please
@tehpiratezorz
2 жыл бұрын
My solution to this problem was that my file path to the input/output layers had a space in it. I created a new filepath without spaces and it was successful.
my classified output is not appearing. What could be the problem?
@jessebuyungo
9 ай бұрын
sorry , what could be the problem
Will it work with landsat 7 data also
@jessebuyungo
9 ай бұрын
yes but at a lwoer resolution
I am having a hard time with this, my results layer ends up flipped like a mirror above my satmap. I think this is making the rest of this project not work.
@jessebuyungo
2 жыл бұрын
Could you please share a snapshot of the output
@jonesderosso
Жыл бұрын
I have the same problem, how can I send you a screenshot?
@simonhanssens8459
8 ай бұрын
is there a solution for this?
Good day Sir, Can I ask the Name of the Plugins used for segmentation? Thank you
@ravindrasingh-bn8uc
Жыл бұрын
orfeo tool box
@jessebuyungo
Жыл бұрын
OTB
@meetgoespro5293
Жыл бұрын
@@jessebuyungo Where do I get this plugin from? I have Orfeo Toolbox provider plugin installed but it's not showing in the panel.
what version are you using?
@jessebuyungo
9 ай бұрын
Hannover
Hi, very informative video but I have this error saying: "Classified_output" has invalid geometry. Please fix the geometry or change the Processing setting to the "Ignore invalid output features" option. This is occured when I try to compute or Dissolve the Classified Output.
@jessebuyungo
2 жыл бұрын
Could you share the error snapshot please
I don't have the segmentation option in my toolbox
@jessebuyungo
Жыл бұрын
Really?
Can I do it in ArcGIS?
@jessebuyungo
2 жыл бұрын
I think you should be able to
you did not mention the source of the imagery
@jessebuyungo
Жыл бұрын
Please feel free to use any high resolution imagery
heey brother l need your help. Isent you an email concerning this topic
@jessebuyungo
4 ай бұрын
please resend the email I have failed to identify it , I am sorry I haven't replied, yet.
@user-nu6xq2nr2b
4 ай бұрын
@@jessebuyungo l managed to figure it out.thank you Jesse
Thanks for this great video! I am trying your method and so far I have no problem from the start until the part of "vector classifier". It keeps showing this: ectorClassifier: itk::ERROR: ListViewParameter(000002495504F300): Value mean_0 not found in the list of choices: Value, Count, NbPixels, Flat, Round, Elong, Perim, meanB1, meanB2, meanB3, stdB1, stdB2, stdB3, MedB1, MedB2, MedB3, VarB1, VarB2, VarB3, KurtB1, KurtB2, KurtB3, SkewB1, SkewB2, SkewB3, NbPixels_1, Flat_1, Round_1, Elong_1, Perim_1, meanB1_1, meanB2_1, meanB3_1, stdB1_1, stdB2_1, stdB3_1, MedB1_1, MedB2_1, MedB3_1, VarB1_1, VarB2_1, VarB3_1, KurtB1_1, KurtB2_1, KurtB3_1, SkewB1_1, SkewB2_1, SkewB3_1.
@jessebuyungo
2 жыл бұрын
Have you computed statistics for the input layer, if Yes make sure the stat layer is joined and the input to the classifier is that joined layer, if you have done this, you can share a snapshot of the error and I see how to help
@user-ys2df3ls7q
2 жыл бұрын
You should try input "meanB1 meanB2" etc. for training vector classifier.
OTB currently support only gdal and ogr provider. Parameter 'inzone.vector.in' uses 'memory' provider Execution failed after 0.04 seconds so, the second half of a video makes no sence to watch
2023-08-25 11:41:41 (WARNING) Segmentation: The input image has no data values but this application does not handle no-data. No-data pixels will be treated as regular pixels. 2023-08-25 11:41:41 (FATAL) Segmentation: itk::ERROR: No OGR driver known to OTB to create and handle a DataSource named . i don't know but my otb doesn't work at all
hi sr. It gives me this error when I apply ''segmetaion': '“.raster.out.tif” files are not supported as outputs for this algorithm' why and how should I correct it? you could tell me. I am looking forward to your response. thank you