Setting Litchi waypoints to take stills

[AMann]

I tried setting up a mission hub project to take many overlapping stills of a field. For Actions at each waypoint, I used "Stay for 5 seconds" to get the drone stabilized before each shot, and then "Take Photo" to get a still shot. During the flight, the drone did not stop at any of the waypoints and only took a single long movie of the entire flight. Where did I go wrong? If you don't mind, please feel free to look at the attached mission file link.

Mission Hub - Litchi

Litchi's Mission Hub lets you plan waypoint flights and later execute them using the mobile Litchi app.

flylitchi.com

I ran a different waypoint mission using DJI's Pilot app and it worked very well, but I'd rather use Litchi if I can get it to work.

Thanks!

 

[sar104]

I've run many similar missions successfully with almost identical settings, so I'm not quite sure why that didn't work. The only differences were that I didn't set a "stay" time at each waypoint (because in photo mode it will stop and stabilize anyway - you don't get a choice), and I didn't start with a "tilt gimbal" command at waypoint #1. Neither of those differences should have led it to take a video. Did you start a video or have the aircraft in video mode before it started the mission, or was the aircraft in photo mode?

If you want to try a profile that definitely works, set the only action at each waypoint to be "Take photo" and make sure that the aircraft is in photo mode before starting.

 

[brett8883]

I tried setting up a mission hub project to take many overlapping stills of a field. For Actions at each waypoint, I used "Stay for 5 seconds" to get the drone stabilized before each shot, and then "Take Photo" to get a still shot. During the flight, the drone did not stop at any of the waypoints and only took a single long movie of the entire flight. Where did I go wrong? If you don't mind, please feel free to look at the attached mission file link.

Mission Hub - Litchi

Litchi's Mission Hub lets you plan waypoint flights and later execute them using the mobile Litchi app.

flylitchi.com

I ran a different waypoint mission using DJI's Pilot app and it worked very well, but I'd rather use Litchi if I can get it to work.

Thanks!

Looks like Sar104 has answered your question but maybe the video isn’t so bad.

You know a video is just series of photos so you could take the video and put it in Premier Pro for example and then export as a JPEG photo sequence exporting every nth frame to a folder and now you have an even higher resolution of information to stitch together to get a higher quality map or whatever you are using it for.

I do this sometimes because it’s easier then setting up up the mapping and takes way less time

 

[GadgetGuy]

Looks like Sar104 has answered your question but maybe the video isn’t so bad.

You know a video is just series of photos so you could take the video and put it in Premier Pro for example and then export as a JPEG photo sequence exporting every nth frame to a folder and now you have an even higher resolution of information to stitch together to get a higher quality map or whatever you are using it for.

I do this sometimes because it’s easier then setting up up the mapping and takes way less time

The video still frame extracts are much lower resolution than an original still photo, so for the same coverage, on a fixed focal length lens, the still photos will always yield higher resolution. Depending upon the resolution needed, the video still frame extracts may be good enough, but you still have to deal with the potential problem of an electronic shutter in motion.

 

[sar104]

Looks like Sar104 has answered your question but maybe the video isn’t so bad.

You know a video is just series of photos so you could take the video and put it in Premier Pro for example and then export as a JPEG photo sequence exporting every nth frame to a folder and now you have an even higher resolution of information to stitch together to get a higher quality map or whatever you are using it for.

I do this sometimes because it’s easier then setting up up the mapping and takes way less time

The problem with that strategy comes if you need to know the exact location of each image - that's difficult to extract from the video. A better, faster strategy if you don't mind photos acquired while moving, is to used the interval timer with a setting that, combined with the aircraft speed, results in suitable overlap. The photo locations are then in the EXIF data.

 

[AMann]

I've run many similar missions successfully with almost identical settings, so I'm not quite sure why that didn't work. The only differences were that I didn't set a "stay" time at each waypoint (because in photo mode it will stop and stabilize anyway - you don't get a choice), and I didn't start with a "tilt gimbal" command at waypoint #1. Neither of those differences should have led it to take a video. Did you start a video or have the aircraft in video mode before it started the mission, or was the aircraft in photo mode?

If you want to try a profile that definitely works, set the only action at each waypoint to be "Take photo" and make sure that the aircraft is in photo mode before starting.

Good points - I can't verify if it was in photo mode before the mission started - will test that using the same mission. I also did not know that it will stop and stabilize prior to taking a photo at a waypoint, thought it would continue flying over. I will also try your suggestion of setting "the only action at each waypoint to be "Take photo" and make sure that the aircraft is in photo mode before starting". Thanks for your time reviewing my settings and responding!

Cheers,

Anthony

 

[brett8883]

The video still frame extracts are much lower resolution than an original still photo, so for the same coverage, on a fixed focal length lens, the still photos will always yield higher resolution. Depending upon the resolution needed, the video still frame extracts may be good enough, but you still have to deal with the potential problem of an electronic shutter in motion.

Yea I meant more coverage. If you use a high enough shutter the motion blur isn't an issue.
When you stitch them together your resulting composite will be higher resolution. The location data is a problem though if you are doing mapping thats certainly true

The problem with that strategy comes if you need to know the exact location of each image - that's difficult to extract from the video. A better, faster strategy if you don't mind photos acquired while moving, is to used the interval timer with a setting that, combined with the aircraft speed, results in suitable overlap. The photo locations are then in the EXIF data.

This is a good idea too!

 

[AMann]

The video still frame extracts are much lower resolution than an original still photo, so for the same coverage, on a fixed focal length lens, the still photos will always yield higher resolution. Depending upon the resolution needed, the video still frame extracts may be good enough, but you still have to deal with the potential problem of an electronic shutter in motion.

for quality sake (I was playing with making maps and 3D models of the site), I was hoping to get RAW stills but it just did video. A flying video with frame grabs would be useful in some situations where that quality or file size is a limitation, but for mapping is would be an issue. Thanks for replying and helping.

 

[AMann]

The problem with that strategy comes if you need to know the exact location of each image - that's difficult to extract from the video. A better, faster strategy if you don't mind photos acquired while moving, is to used the interval timer with a setting that, combined with the aircraft speed, results in suitable overlap. The photo locations are then in the EXIF data.

This is the method they used at Amboy Crater, CA when looking for remains of a couple that were lost in a lava field that succumbed to the Mojave summer heat. It was too hot for search dogs and the searchers, so they used drones and covered the area flying it during the hot part of the day and scanned the high resolution photos to look for them.

 

[brett8883]

This is the method they used at Amboy Crater, CA when looking for remains of a couple that were lost in a lava field that succumbed to the Mojave summer heat. It was too hot for search dogs and the searchers, so they used drones and covered the area flying it during the hot part of the day and scanned the high resolution photos to look for them.

Yea I've done a few of those. The trick is to fly as low as possible and get the correct angle to the earth but what program are you using that allows you to get a composite 3D DNG? I didn't think a raw file could exist in a 3D space.

 

[AMann]

Yea I've done a few of those. The trick is to fly as low as possible and get the correct angle to the earth but what program are you using that allows you to get a composite 3D DNG? I didn't think a raw file could exist in a 3D space.

I'm using Agisoft Photoscan. This is from 100 RAW stills I got when using waypoint mode in DJI Pilot while flying at the same site as the Litchi mission above. The product isn't a DNG, I meant just the source files were.

 

[sar104]

This is the method they used at Amboy Crater, CA when looking for remains of a couple that were lost in a lava field that succumbed to the Mojave summer heat. It was too hot for search dogs and the searchers, so they used drones and covered the area flying it during the hot part of the day and scanned the high resolution photos to look for them.

That's the method I use for search and rescue too. Even when conditions are appropriate for ground search, looking from above still adds another significant element to the overall POD.

 

[brett8883]

I'm using Agisoft Photoscan. This is from 100 RAW stills I got when using waypoint mode in DJI Pilot while flying at the same site as the Litchi mission above. The product isn't a DNG, I meant just the source files were.
View attachment 71979

So you think it’s better results using the raw files? Reason I ask is I’ve just been doing it with JPEGs because I figure unless I process them first the software is gonna convert them to JPEG or TIFF before it can stitch them anyway and seems like a lot less work and processing time.

But if you think it’s better I’ll give it a try

 

[AMann]

So you think it’s better results using the raw files? Reason I ask is I’ve just been doing it with JPEGs because I figure unless I process them first the software is gonna convert them to JPEG or TIFF before it can stitch them anyway and seems like a lot less work and processing time.

But if you think it’s better I’ll give it a try

With RAW photos you can post process them for better image clarity, sharpness, exposure, dynamic range and color. If you shoot with JPG's, you cant post process them as much. I shoot most everything in RAW, but when doing shots when I am in a hurry and don't want to keep them, I will use JPG's also.

 

[AMann]

So you think it’s better results using the raw files? Reason I ask is I’ve just been doing it with JPEGs because I figure unless I process them first the software is gonna convert them to JPEG or TIFF before it can stitch them anyway and seems like a lot less work and processing time.

But if you think it’s better I’ll give it a try

With RAW photos you can post process them as needed for better image clarity, sharpness, exposure, dynamic range and color. If you shoot with JPG's, you cant post process them as much. I shoot most everything in RAW, but when doing shots when I am in a hurry and don't want to keep them, I will use JPG's also.

 

[brett8883]

With RAW photos you can post process them for better image clarity, sharpness, exposure, dynamic range and color. If you shoot with JPG's, you cant post process them as much. I shoot most everything in RAW, but when doing shots when I am in a hurry and don't want to keep them, I will use JPG's also.

I know but I never considered post processing files before putting them into mapping software. What do you do to them before hand? Do you batch process then or go through them one by one?

 

[AMann]

Another question - is there a way to automatically get a waypoint grid in Litchi like you can in DJI Pilot? I had to hand draw my grid in that Litchi mission I posted above, but DJI Pilot does it for me after I supply it with just the camera sensor and lens specs and the four corners of the flight area.

 

[sar104]

Another question - is there a way to automatically get a waypoint grid in Litchi like you can in DJI Pilot? I had to hand draw my grid in that Litchi mission I posted above, but DJI Pilot does it for me after I supply it with just the camera sensor and lens specs and the four corners of the flight area.

As far as I can tell - not unless you write your own program to create grids. That's what I ended up doing.

 

[AMann]

As far as I can tell - not unless you write your own program to create grids. That's what I ended up doing.

View attachment 71989

Thanks. You just gave me a higher wattage lightbulb moment- perhaps I can set it up in DJI pilot, then export the CSV file into Litchi.

 

[sar104]

Thanks. You just gave me a higher wattage lightbulb moment- perhaps I can set it up in DJI pilot, then export the CSV file into Litchi.

Maybe. The Litchi profile CSV has a lot of fields containing waypoint information, but you might be able to import the basic lat/long/height file and then add the other stuff in the Litchi mission hub.