Hi, I don't know if you believe me but I watched all your videos related to Agisoft and learned a lot so for these series of videos I am very grateful. I was trying to follow you step by step until I needed to draw a new polyline over the old one to divide the first shape, but the old shape vertex markers and points were not visible to snap on. I tried a lot to find out why but with no result. I am using the latest version of Agisoft.
Thanks for the instructional videos. I want to ask a question. In a documentary about Egypt, they were measuring the length and height of the pyramids using Realty Capture. How to make basic measurements such as door, window lengths or building height on the building where we created the 3d mesh model with Agisoft and pix4d? It would be very helpful for survey areas if you prepared a video on this subject. thanks.
Hi - thanks for watching and for your feedback. I have a bit of time coming up, so I will make a video along these lines. I will let you know once it's up and ready!
I'm working on a video to address this, but I would suggest editing colour, contrast, saturation etc. before using Metashape. It is always better to adjust images as early as possible as it helps all of the following processes.
@@geospatialtips I adjusted all the things however the different dates flying there is seasonal variation that time both flight overlap area mask has came. We can't adjust manually.
@@MrABN143 Thanks, understood. In that case, I'd suggest you look at an alternative software for your orthomosaic. Metashape is not designed to handle seasonal variability. I use Inpho Orthovista and enjoy it's capabilities. Bingo is also very good I believe. But, if you like, send me a screenshot of your orthomosaic to geospatialtips@gmail.com and I will take a look and perhaps offer some other suggestions?
Hi TT Thanks for watching. This particular system isn't all that great, quite old now if I think about it, but a decent GPU helps. Intel i7 8700k 64GB RAM Nvidia 3080 GPU Win 10 Pro Processing drive is a 1TB Samsung EVO NVME
While watching this video again after watching the other ones I see that you didn't classify ground points and built the DEM from the dense cloud using all classes 2:01 When measuring stockpiles is it typical to do it this way or are there times when we should classify ground points and build the DEM as a DTM, using the ground class. Is that how we remove conveyors and other machinery from the piles? THANKS! Edit: When I did classify ground points I noticed that the top of one pile disappeared. I did it three times with different settings but always lost the top of the pile. I saw this when I filtered by ground class. I don't think this is good since I need to calculate the volume of the pile. I'm confused now...
Hi - you are correct that if you have any machinery, conveyors or any other clutter you should cleanup the dense cloud and ensure that the points you are creating your DTM from are only the valid ground points. If the automated ground classification is failing (possibly due to a steep incline) you can manually classify those points at the top of the pile to the ground class. To do this, select the points in question with the selection tool and the press "Ctrl+Shift+C" or go "Tools-Dense Cloud-Assign Class" and say assign "From any class" to "Ground". That will add those points to the ground class as well and then you can continue to create your clean and accurate DTM from the ground class only. Hope that helps!
@@geospatialtips That helps! There are many videos and tutorials online but yours are the most helpful. Then when you respond to questions in the comments it's even better 🙂 I'll have more questions. I hope you can continue to help everyone who sees these comments. Thanks!
I'm back again LOL When you calculated the different results with the two different methods, how did you decide that the 2-part method gave the correct answer? Was it just that the profile(s) were more uniform? How confident are you that the result was correct and why? I'm asking the tough questions because I want to know how to answer them for myself and potential clients. I don't feel right about selling a service when I lack confidence in the results and/or wouldn't be able to answer that question for a client. I've been practicing a lot and even did some of my Metashape projects on the Measure Ground Control platform to compare the results. MGC uses Pix4D cloud for processing. The results are similar between MGC and Metashape but both seem very large. For example one pile came out to 600 yd3. I started thinking about how many loads of material that is. With big 10yd trucks there would be 60 truck loads of material. It just seems like too much. Is it normal for the actual volume to be different that what it appears to be visually? Thanks again!!
Hi again! Looking at that shape, where some of the stockpile has been dumped against a slope, it is immediately apparent that a single volume won't work because we don't have a base file to calculate volume above and we are only able to use planes in Metashape, hence a minimum of two planes are needed to accurately portray the base in this instance. Once those two shapes had been placed, then yes I was much happier with the profile I saw as it was more uniform and thus the volume result would be more reliable. Because I have considered the shape and applied logic to determine how many regions would define it, added to the fact that the profiles were better in the second instance and, if I am honest, just through experience, I am confident that the two part volume was the correct one. It's good that you are asking these questions, rather than just blindly accepting what Metashape gives you. A nice idea to compare software packages as well, that further affirms your results. As for whether one can be deceived by a volume that looks small, but you get a large value, yes that does happen fairly often. When in doubt, go back to basic maths. Volume = length x breadth x height. If it gives you an answer close to what your volume is telling you, then it must be right. However, if for whatever reason your scale is off, that could mess things up badly, but I think you have control, so that should address that type of problem. I'd be happy to run the data myself and compare for you, please just email me the link and I'll do some processing. geospatialtips@gmail.com Cheers
@@geospatialtips Thanks for explaining and thanks for offering to run the data. I'll upload the photos tonight and send the link. You are really the only person I've come across who offers their time like this. It's a great opportunity for me along with those who watch your videos and read these comments 🙂
@@jerseyshoredroneservices225 Thanks for your kind comments. I'm really just happy when I can help out and maybe it goes some way to improving the industry as a whole, even if just the smallest little bit. I realise this video may have asked more questions than it has answered, so I took a bit of time today to record a video that looks at this exact stockpile and answers why the two part approach is good and I compare the calculations in two other software packages. I will have it uploaded as soon as I can and then you'll see the logic I have applied. I'm looking forward to receiving your dataset as well and working through it with you!
Im having trouble understanding best fit and mean plane. I am trying to calculate the volume of a pile of mulch sitting on a slope. Would best fit be the correct option for this scenario?
Hi Dennis - you are correct. In this case the best fit makes the most sense (unless you had a custom base from before the pile was there). The mean plane would mean you lose a bit of material and the volume you get would be off. I'm not sure if you have watched this video, but I explain the differences between each plane type in a little more depth, maybe it'll help you. ua-cam.com/video/9t-EJBVqgLE/v-deo.html
Thanks for watching! You are correct that I didn't classify the points in this case and likely you should be doing that if there is any risk of machinery or clutter on the stockpiles.
Hi - you can clip the dense cloud by using the resize region option. You will see me doing this at 57seconds into the video. The tool can be found in the toolbar at the top of the Metashape window. Hope that helps!
I'm just starting to do stock piles so this is very helpful. Unfortunately I have several piles in bunkers and some up against vertical walls on one side. There are also some where the edged of the piles are obscured by trees. Some piles have conveyors that need to be removed. I've been making Ortho's and DTM's but now I think I need to watch your other video to see how you got to where this video started. I think I've been doing it differently. Do you offer courses that go into more detail than the YT videos? Thanks!
Hi - yes, there shouldn't be any issue with that. You can import images, point clouds, DTMs, DSMs etc. into Metashape and then work with them, manipulate or edit them as needed. Anything you import will need its relevant metadata, so be sure to export that from Dronelink or include it in your settings and the rest should be easy enough. Of course you could just use Dronelink for the capture and then do all of the data processing in Metashape. The choice is yours.
@@geospatialtips Thanks. Yeah its kinda the latter that I was thinking of. New to all this so just finding my way round some of the software and terminology etc. Thanks
Sure thing - Metashape will accept almost any from of imagery, it's very forgiving. You'll want to make sure it imports the metadata such as GPS tags and also orientation data and accuracy info (if using an RTK drone). If it should be importing that info and it isn't doing so, you can enable the setting in Tools/Preferences/Advanced.
Hi - effectively you apply the same principle, but in Metashape you take the "Volume Below" value. So, it is the inverse volume you calculate. Create you model and draw the boundary around your excavated region, using a surface that is as flat as possible. Then compute the volume and volume below should be the excavated value. If the volume is more complex, you may need a base dataset to subtract from. If you need, send me some details in an email and I'll be happy to help you where I can. geospatialtips@gmail.com
Please consider hitting the subscribe button 🙏 It helps me out SO MUCH and will allow me to create more helpful videos! 😊
This gent is an absolute wizard at all things geospatial. Loving the channel, and can't wait to see what vid drops next.
Thks for this free lesson, from Bazil! Keep goin to your channel!
Hi, I don't know if you believe me but I watched all your videos related to Agisoft and learned a lot so for these series of videos I am very grateful. I was trying to follow you step by step until I needed to draw a new polyline over the old one to divide the first shape, but the old shape vertex markers and points were not visible to snap on. I tried a lot to find out why but with no result. I am using the latest version of Agisoft.
i absolutely love your videos you have taught me so much
Glad you like them!
great. more real life work tips would be welcomed. thnx
More to come! Thanks
Thanks for the instructional videos. I want to ask a question. In a documentary about Egypt, they were measuring the length and height of the pyramids using Realty Capture. How to make basic measurements such as door, window lengths or building height on the building where we created the 3d mesh model with Agisoft and pix4d? It would be very helpful for survey areas if you prepared a video on this subject. thanks.
Hi - thanks for watching and for your feedback.
I have a bit of time coming up, so I will make a video along these lines. I will let you know once it's up and ready!
Can you make a video on ortho colour adjust for particular areas
I'm working on a video to address this, but I would suggest editing colour, contrast, saturation etc. before using Metashape. It is always better to adjust images as early as possible as it helps all of the following processes.
@@geospatialtips I adjusted all the things however the different dates flying there is seasonal variation that time both flight overlap area mask has came. We can't adjust manually.
@@MrABN143 Thanks, understood. In that case, I'd suggest you look at an alternative software for your orthomosaic. Metashape is not designed to handle seasonal variability.
I use Inpho Orthovista and enjoy it's capabilities. Bingo is also very good I believe.
But, if you like, send me a screenshot of your orthomosaic to geospatialtips@gmail.com and I will take a look and perhaps offer some other suggestions?
@@geospatialtips sure, I will share.
Very interesting and helpful Video.
What are the specifics of your hardware you are using for agisoft?
Can't wait to see more of your Videos :)
Hi TT
Thanks for watching.
This particular system isn't all that great, quite old now if I think about it, but a decent GPU helps.
Intel i7 8700k
64GB RAM
Nvidia 3080 GPU
Win 10 Pro
Processing drive is a 1TB Samsung EVO NVME
Nice tutorial
While watching this video again after watching the other ones I see that you didn't classify ground points and built the DEM from the dense cloud using all classes 2:01
When measuring stockpiles is it typical to do it this way or are there times when we should classify ground points and build the DEM as a DTM, using the ground class. Is that how we remove conveyors and other machinery from the piles? THANKS!
Edit: When I did classify ground points I noticed that the top of one pile disappeared. I did it three times with different settings but always lost the top of the pile. I saw this when I filtered by ground class. I don't think this is good since I need to calculate the volume of the pile.
I'm confused now...
Hi - you are correct that if you have any machinery, conveyors or any other clutter you should cleanup the dense cloud and ensure that the points you are creating your DTM from are only the valid ground points.
If the automated ground classification is failing (possibly due to a steep incline) you can manually classify those points at the top of the pile to the ground class. To do this, select the points in question with the selection tool and the press "Ctrl+Shift+C" or go "Tools-Dense Cloud-Assign Class" and say assign "From any class" to "Ground".
That will add those points to the ground class as well and then you can continue to create your clean and accurate DTM from the ground class only.
Hope that helps!
@@geospatialtips
That helps!
There are many videos and tutorials online but yours are the most helpful. Then when you respond to questions in the comments it's even better 🙂
I'll have more questions. I hope you can continue to help everyone who sees these comments. Thanks!
I'm back again LOL
When you calculated the different results with the two different methods, how did you decide that the 2-part method gave the correct answer? Was it just that the profile(s) were more uniform? How confident are you that the result was correct and why?
I'm asking the tough questions because I want to know how to answer them for myself and potential clients. I don't feel right about selling a service when I lack confidence in the results and/or wouldn't be able to answer that question for a client.
I've been practicing a lot and even did some of my Metashape projects on the Measure Ground Control platform to compare the results. MGC uses Pix4D cloud for processing.
The results are similar between MGC and Metashape but both seem very large. For example one pile came out to 600 yd3. I started thinking about how many loads of material that is. With big 10yd trucks there would be 60 truck loads of material. It just seems like too much. Is it normal for the actual volume to be different that what it appears to be visually?
Thanks again!!
Hi again!
Looking at that shape, where some of the stockpile has been dumped against a slope, it is immediately apparent that a single volume won't work because we don't have a base file to calculate volume above and we are only able to use planes in Metashape, hence a minimum of two planes are needed to accurately portray the base in this instance.
Once those two shapes had been placed, then yes I was much happier with the profile I saw as it was more uniform and thus the volume result would be more reliable. Because I have considered the shape and applied logic to determine how many regions would define it, added to the fact that the profiles were better in the second instance and, if I am honest, just through experience, I am confident that the two part volume was the correct one.
It's good that you are asking these questions, rather than just blindly accepting what Metashape gives you. A nice idea to compare software packages as well, that further affirms your results.
As for whether one can be deceived by a volume that looks small, but you get a large value, yes that does happen fairly often. When in doubt, go back to basic maths. Volume = length x breadth x height. If it gives you an answer close to what your volume is telling you, then it must be right. However, if for whatever reason your scale is off, that could mess things up badly, but I think you have control, so that should address that type of problem.
I'd be happy to run the data myself and compare for you, please just email me the link and I'll do some processing.
geospatialtips@gmail.com
Cheers
@@geospatialtips
Thanks for explaining and thanks for offering to run the data. I'll upload the photos tonight and send the link.
You are really the only person I've come across who offers their time like this. It's a great opportunity for me along with those who watch your videos and read these comments 🙂
@@jerseyshoredroneservices225
Thanks for your kind comments. I'm really just happy when I can help out and maybe it goes some way to improving the industry as a whole, even if just the smallest little bit.
I realise this video may have asked more questions than it has answered, so I took a bit of time today to record a video that looks at this exact stockpile and answers why the two part approach is good and I compare the calculations in two other software packages.
I will have it uploaded as soon as I can and then you'll see the logic I have applied.
I'm looking forward to receiving your dataset as well and working through it with you!
Im having trouble understanding best fit and mean plane. I am trying to calculate the volume of a pile of mulch sitting on a slope. Would best fit be the correct option for this scenario?
Hi Dennis - you are correct. In this case the best fit makes the most sense (unless you had a custom base from before the pile was there). The mean plane would mean you lose a bit of material and the volume you get would be off. I'm not sure if you have watched this video, but I explain the differences between each plane type in a little more depth, maybe it'll help you. ua-cam.com/video/9t-EJBVqgLE/v-deo.html
Ganhou uma inscrita! Excelente explicação! Uma dúvida: não houve classificação de pontos, então esse MDE é o MDS?
Thanks for watching!
You are correct that I didn't classify the points in this case and likely you should be doing that if there is any risk of machinery or clutter on the stockpiles.
Nice darling ❤️
thanks
Hello
How did you clip the densecloud at the size ouf your boundery box and generate a DEM only the size of the box ( not all the photo set )
Hi - you can clip the dense cloud by using the resize region option. You will see me doing this at 57seconds into the video. The tool can be found in the toolbar at the top of the Metashape window. Hope that helps!
I'm just starting to do stock piles so this is very helpful. Unfortunately I have several piles in bunkers and some up against vertical walls on one side. There are also some where the edged of the piles are obscured by trees. Some piles have conveyors that need to be removed.
I've been making Ortho's and DTM's but now I think I need to watch your other video to see how you got to where this video started. I think I've been doing it differently.
Do you offer courses that go into more detail than the YT videos?
Thanks!
Hi there.
Would you like to mail me at geospatialtips@gmail.com and I can share more info and we can figure out how to help you with those volumes?
@@geospatialtips
I will. Thank you!
@@geospatialtips
I just sent the email. It's from Jerseyshore.....
Thank you!
If I run my flight and data capture using dronelink can I then upload the data into Metashape?
Hi - yes, there shouldn't be any issue with that.
You can import images, point clouds, DTMs, DSMs etc. into Metashape and then work with them, manipulate or edit them as needed.
Anything you import will need its relevant metadata, so be sure to export that from Dronelink or include it in your settings and the rest should be easy enough.
Of course you could just use Dronelink for the capture and then do all of the data processing in Metashape. The choice is yours.
@@geospatialtips Thanks. Yeah its kinda the latter that I was thinking of. New to all this so just finding my way round some of the software and terminology etc. Thanks
Sure thing - Metashape will accept almost any from of imagery, it's very forgiving. You'll want to make sure it imports the metadata such as GPS tags and also orientation data and accuracy info (if using an RTK drone). If it should be importing that info and it isn't doing so, you can enable the setting in Tools/Preferences/Advanced.
@@geospatialtips Thanks much appreciated
How can I calculate the volume excavated? .. that does not protrude from the ground
Hi - effectively you apply the same principle, but in Metashape you take the "Volume Below" value. So, it is the inverse volume you calculate.
Create you model and draw the boundary around your excavated region, using a surface that is as flat as possible. Then compute the volume and volume below should be the excavated value.
If the volume is more complex, you may need a base dataset to subtract from. If you need, send me some details in an email and I'll be happy to help you where I can. geospatialtips@gmail.com
Please upload basic to professional tutorial videos step to step