I got one tips for, been working a lot with photography before I started with surveying. A dead highlight zone is always dead. But a dark underexposed photo still got all the info. It’s better to letting the camera shoot 1 or 2 steps low on exposure. The survey will get a higher quality and accuracy but not looking as nice for the naked eye. You can do the same process but boosting exposure to make them brighter. And don’t get blown out highlights Cheers from Sweden 😊
Hi Paul - no, I can't say that I have. But it sounds like something worth a try. So long as the corrections are some kind of average across all of the images, then great. I would be concerned if it is adjusting each image based on it's own data, but if you are getting consistent exposure etc. across all the images, excellent.
Is there a particular reason for you using a different software other than Agisoft metashape in this video for showing the point cloud part? And Thank you for sharing this amazing content, please don't stop, you are helping me out tremendously, and probably other pilots as well. With much appreciation: A beginner industrial drone pilot
Hey Tamas - thanks for watching, I appreciate it and your comment. I used Global Mapper to display things in the video just because it was easier for me to load two different point clouds side-by-side, draw cross sections, evaluate the imagery at the same time etc. Everything was still produced using Metashape using the same workflow I always do. But the point is really that this little tip will make a big difference regardless of what photogrammetry software you use.
Thanks for the comment! Nice idea for the video...I just don't have a copy of Lightroom at the moment, but in time I think there is value in such a comparison.
William Faucher recommends to maximize the shadows while minimizing the highlights (in his video: "RealityCapture to UE5 - Workflow Tutorial"). Why do you differ on this? Is this something that is dependent on your data?
Hi - yes, I think it is very dependent on the data you are working with. I'd suggest that it is best to know what tools are available and what each would achieve and then make a call based on your specific situation, needs and data. I don't believe there is a magic bullet solution for every case. Hope that helps a little.
do you know how to edit the mosaic instead of each raw photos? I'm talking about 28k photos, 22k Ha of land, 60 flights, and various lightning condition for each flight. it will be easier to work with the mosaic to mask the clouds and enhance the color when youre working with big dataset
You could Photoshop it all, but that's a bit of a pain to do. Realistically, with a project of the size you have mentioned, I would create the mosaic in a specialized software, such as OrthoVista. (Bingo would do much the same) In OV you can set it to balance data per flight and across multiple days, edit the radiometry of the images individually, within a region or per flight and so on. Then, create the mosaic which would thus be much better from the start. Any remaining edits for colour, seamlines, balancing etc. can be reviewed and edited and then the mosaic, or just the are you need, can be redone.
@@geospatialtips thanks or the reply. Never heard those software, but ill look it up. Currently im using photoshop to edit the tiled mosaic. The fact that PS cant save georeference information was solved by exporting the tiled mosaic together with the World file (.jgw). Ive never found geospatial software that can edit raster data just like PS did. So the method above can be a workaround
@@mikealpha4462 Yes, the loss of referencing is a pain, but as you say, it can be overcome. OrthoVista is pretty expensive, owned and distributed by Trimble these days, but it certainly gets the job done. You would export rectified images from Metashape (or whichever program you use) import and edit, then create the mosaic in OV.
Any tips for preserving orientation data on all images processed through Rawtherapee? Processing the same set of photos in Pix4D, one set adjusted with Rawtherapee, I am finding that the adjusted photo set has had orientations stripped from some of the photos. This is flagged in Pix4D when trying to select Acurate Geolocation and Orientation for Calibration Method, and Omega, Phi and Kappa are all unpopulated.
Hi, thanks for watching. Off hand, I’m not sure why it has stripped them out, very odd and I haven’t seen it before. I would think there may be a setting where we select “Keep all metadata” or “export attributes” or something similar. (Sorry, not close to a PC to confirm right now) If all else fails, once Rawtherapee is done, copy the orientation metadata into those images from the original images. You could use a tool like Exiftool to do that. It’s free to use and there is a lot of info online about it. Else, in Pix4D, load the original images, export those orientations to a file, then import them with the Rawtherapee images. Both options are the long way around, but will work if you can’t find a setting in Rawtherapee to fix it.
there is an option on rawtherapee on metadata the says "META", right next to "exposure", you need to check all the boxes so the images can bring these info, including gps tags
Thanks Harry! I’ll be honest and say I’ve never seen any of the courses that are out there myself, so I couldn’t recommend one over the other. I see there are many if you Google it, lots of them with very good reviews. That might be a good starting point. I’ve been thinking of creating one myself, but to do it properly will take a lot of time and I’m not yet sure the demand justifies that. I hope you find a good solution and would be keen to hear which you choose.
You’re absolutely correct, doing things properly at the time of capture is always best. But sometimes, it isn’t possible or doesn’t work out how we plan. Or, as is my case where I process data for clients, I have no control over the capture and work with what the client sends me. I find that the adjustments mentioned here often help to improve the input data. It’s always good to have options and ideas to improve our datasets.
Hi First, try the approach in this video, which I imagine you may now have done. Then, run your alignment with more tie points and run your mesh/model with higher settings. That should help a bit. If you still don’t have success, you may need to capture more images and various angles, and remember to try and do this in overcast conditions which will give better results.
Hi - they’re all maintained within the file exif data on export. There won’t be any loss. If you use something like Lightroom you have the option to keep all, change it, keep some or remove all. So you won’t have to worry about losing that component of your data, regardless of which software you use.
I've been doing a similar process utilizing Adobe bridge and camera raw. Open bridge to the folder with the images, right click on one image and select open in camera raw. Make the adjustments and close the image. You'll go back to Bridge and see your adjustments in the image. To apply the same adjustments to another image or any selection of images, right click on the one you just did and select develop settings, copy camera raw settings. Now choose another photo, or all of them and paste the camera raw settings. When I first did this I noticed that the adjustments were not reflected when looking at the photos outside of Adobe, for example in the windows image viewer. To fix that I used Photoshop... File, scripts, image processor. That tool allows you to re-save a whole directory of images to another folder. If you choose to save in the same folder it will create a subfolder for the images. The images that it saves to this new folder have the adjustments baked-in. Now I'm curious to see if this Google tool will do better than Adobe...
Great feedback, thanks! There are a few ways to do this and I would think numerous pieces of software that do pretty much the same thing. In the Adobe suite, I use Lightroom quite a lot as I can make all of the adjustments and exports in one software and with just a few clicks. I this video, my aim as to see if I could achieve the same thing with a free and opensource piece of software. Turns out it is possible :)
i just did it and it kept some information in the exif, not all, it kept the coordinates for example, but it didn't keep the precision values, it's something to take into account if you use metashape and you take the presicion values from the images
I got one tips for, been working a lot with photography before I started with surveying. A dead highlight zone is always dead. But a dark underexposed photo still got all the info. It’s better to letting the camera shoot 1 or 2 steps low on exposure. The survey will get a higher quality and accuracy but not looking as nice for the naked eye. You can do the same process but boosting exposure to make them brighter. And don’t get blown out highlights
Cheers from Sweden 😊
Thanks for the tip!
Yes, it is always best to plan well before you fly a project because that will lead to the best possible result.
@@Cherbanov that's absolutely not 100% true.
Oh my god never thinked about colorgrading and highlights and shadow, you are right thank you
As always great job
Great video as usual, I’ll reach out tomorrow afternoon and hopefully we can schedule that talk for next week
Thank you so much. This made an incredible difference to my washed out photos!
Fantastic!
Wow, that's a great video, thanks for sharing. Amazing tool.
Thank you for sharing how to. I was using Darktable but this seems simpler to batch and just as good results!
Have you tried using literoom and applying the auto settings to all of the images to get consistent exposure etc across all images?
Hi Paul - no, I can't say that I have. But it sounds like something worth a try. So long as the corrections are some kind of average across all of the images, then great. I would be concerned if it is adjusting each image based on it's own data, but if you are getting consistent exposure etc. across all the images, excellent.
@geospatialtips I've tried it after asking this and it turns out pretty good.
Is there a particular reason for you using a different software other than Agisoft metashape in this video for showing the point cloud part?
And Thank you for sharing this amazing content, please don't stop, you are helping me out tremendously, and probably other pilots as well.
With much appreciation: A beginner industrial drone pilot
Hey Tamas - thanks for watching, I appreciate it and your comment.
I used Global Mapper to display things in the video just because it was easier for me to load two different point clouds side-by-side, draw cross sections, evaluate the imagery at the same time etc. Everything was still produced using Metashape using the same workflow I always do. But the point is really that this little tip will make a big difference regardless of what photogrammetry software you use.
Great video! I would be very interested in a Lightroom video, using both RAW and JPEG and showing the difference in the results.
Thanks for the comment!
Nice idea for the video...I just don't have a copy of Lightroom at the moment, but in time I think there is value in such a comparison.
Thank You!! This was so helpful, you've earned a sub
William Faucher recommends to maximize the shadows while minimizing the highlights (in his video: "RealityCapture to UE5 - Workflow Tutorial"). Why do you differ on this? Is this something that is dependent on your data?
Hi - yes, I think it is very dependent on the data you are working with.
I'd suggest that it is best to know what tools are available and what each would achieve and then make a call based on your specific situation, needs and data. I don't believe there is a magic bullet solution for every case. Hope that helps a little.
do you know how to edit the mosaic instead of each raw photos? I'm talking about 28k photos, 22k Ha of land, 60 flights, and various lightning condition for each flight. it will be easier to work with the mosaic to mask the clouds and enhance the color when youre working with big dataset
You could Photoshop it all, but that's a bit of a pain to do.
Realistically, with a project of the size you have mentioned, I would create the mosaic in a specialized software, such as OrthoVista. (Bingo would do much the same) In OV you can set it to balance data per flight and across multiple days, edit the radiometry of the images individually, within a region or per flight and so on. Then, create the mosaic which would thus be much better from the start. Any remaining edits for colour, seamlines, balancing etc. can be reviewed and edited and then the mosaic, or just the are you need, can be redone.
@@geospatialtips thanks or the reply. Never heard those software, but ill look it up. Currently im using photoshop to edit the tiled mosaic. The fact that PS cant save georeference information was solved by exporting the tiled mosaic together with the World file (.jgw). Ive never found geospatial software that can edit raster data just like PS did. So the method above can be a workaround
@@mikealpha4462 Yes, the loss of referencing is a pain, but as you say, it can be overcome.
OrthoVista is pretty expensive, owned and distributed by Trimble these days, but it certainly gets the job done. You would export rectified images from Metashape (or whichever program you use) import and edit, then create the mosaic in OV.
Any tips for preserving orientation data on all images processed through Rawtherapee? Processing the same set of photos in Pix4D, one set adjusted with Rawtherapee, I am finding that the adjusted photo set has had orientations stripped from some of the photos. This is flagged in Pix4D when trying to select Acurate Geolocation and Orientation for Calibration Method, and Omega, Phi and Kappa are all unpopulated.
Hi, thanks for watching.
Off hand, I’m not sure why it has stripped them out, very odd and I haven’t seen it before. I would think there may be a setting where we select “Keep all metadata” or “export attributes” or something similar. (Sorry, not close to a PC to confirm right now)
If all else fails, once Rawtherapee is done, copy the orientation metadata into those images from the original images.
You could use a tool like Exiftool to do that. It’s free to use and there is a lot of info online about it.
Else, in Pix4D, load the original images, export those orientations to a file, then import them with the Rawtherapee images.
Both options are the long way around, but will work if you can’t find a setting in Rawtherapee to fix it.
there is an option on rawtherapee on metadata the says "META", right next to "exposure", you need to check all the boxes so the images can bring these info, including gps tags
Awesome, thank you for this!
You're most welcome. Thanks for watching!
Amazing, could you please recommend me a photogrammetry course!?
Thanks Harry!
I’ll be honest and say I’ve never seen any of the courses that are out there myself, so I couldn’t recommend one over the other. I see there are many if you Google it, lots of them with very good reviews. That might be a good starting point.
I’ve been thinking of creating one myself, but to do it properly will take a lot of time and I’m not yet sure the demand justifies that.
I hope you find a good solution and would be keen to hear which you choose.
Use ND filters. And try auto mode on the drone. The more you do practically , the less you do in post. I learned the hard one on the drone
You’re absolutely correct, doing things properly at the time of capture is always best.
But sometimes, it isn’t possible or doesn’t work out how we plan.
Or, as is my case where I process data for clients, I have no control over the capture and work with what the client sends me. I find that the adjustments mentioned here often help to improve the input data.
It’s always good to have options and ideas to improve our datasets.
@@geospatialtips Agreed!! :)
hai, how to fix hole in agisoft? becaune my 3d house get some holes
Hi
First, try the approach in this video, which I imagine you may now have done. Then, run your alignment with more tie points and run your mesh/model with higher settings. That should help a bit.
If you still don’t have success, you may need to capture more images and various angles, and remember to try and do this in overcast conditions which will give better results.
My text appears weird when I open the software.
What about metadata? GPS coordinates? RTK corrections?
Hi - they’re all maintained within the file exif data on export. There won’t be any loss. If you use something like Lightroom you have the option to keep all, change it, keep some or remove all. So you won’t have to worry about losing that component of your data, regardless of which software you use.
I've been doing a similar process utilizing Adobe bridge and camera raw. Open bridge to the folder with the images, right click on one image and select open in camera raw. Make the adjustments and close the image. You'll go back to Bridge and see your adjustments in the image.
To apply the same adjustments to another image or any selection of images, right click on the one you just did and select develop settings, copy camera raw settings.
Now choose another photo, or all of them and paste the camera raw settings.
When I first did this I noticed that the adjustments were not reflected when looking at the photos outside of Adobe, for example in the windows image viewer. To fix that I used Photoshop... File, scripts, image processor. That tool allows you to re-save a whole directory of images to another folder. If you choose to save in the same folder it will create a subfolder for the images. The images that it saves to this new folder have the adjustments baked-in.
Now I'm curious to see if this Google tool will do better than Adobe...
Great feedback, thanks!
There are a few ways to do this and I would think numerous pieces of software that do pretty much the same thing.
In the Adobe suite, I use Lightroom quite a lot as I can make all of the adjustments and exports in one software and with just a few clicks.
I this video, my aim as to see if I could achieve the same thing with a free and opensource piece of software. Turns out it is possible :)
@@geospatialtips
Thank you for all the great education on your channel and also making me want to learn Global Mapper!
i just did it and it kept some information in the exif, not all, it kept the coordinates for example, but it didn't keep the precision values, it's something to take into account if you use metashape and you take the presicion values from the images
doesnt adobe do all this as well??
Correct! I know Adobe Lightroom can do it basically the same way, in a batch.
This is just a free option for those folks that don't have Adobe.
Wow
Mmmmm so good
Thanks!