Just curious, when a depth sensor is in use as a camera, there are these black balls the fly through the frame, different sizes and speeds. In your example footage you see a few of these. They don't appear to be typical noise, so what are these black balls?
Hello, what sensor would you recommend for 3d scanning boats. I have a friend that uses the Intel realsence, but I wonder if there is a better alternative out there in its price range.
Thanks for the great video. How would it perform if it takes a point cloud of an object high contrast? Like highly reflective ropes wrapping around a black object.
Hey, thank you for this, I was thinking about getting the Azure exactly for depth maps, stationary camera but moving circus people. Now I'm not so certain. Can you recommend something better in a similar price range? Or am I looking for more expensive equipment?
Hi, thanks for the review, this was very helpful! I would love to watch a comparison of all the sensors you mentioned, the robotics club at my school needs to purchase a good depth sensor but we aren't sure which one to buy. Thanks again, loved this vid!
Really interesting video! Especially the part on reconstruction artifices. I was hoping to develop an application that can segment coloured cables in a scene with the RGB sensor. Then extract depth information to construct a spline for each cable. Do you think this parallax effect would cause many issues?
Glad it was helpful! This depends heavily on the context of the application. Bare in mind that the sensors (IR and RGB) are only horizontally displaced. So if you had the cables placed horizontally and separately vertically, you might not see this parallax effect at all. But if you place the cables vertically, this parallax effect will definitely impact the depth sensing, especially if the cables are thin.
@@robotfpv Thanks! Yea, unfortunately the cables are 3 to 1mm in diameter. However, they will definitely be placed horizontally and separated vertically. They are just placed on a table so I can mount the Kinect anywhere and control lighting. I am wondering if capturing depth information on a 1mm diameter cable is unrealistic with this sensor.
@@Fergus858 1mm will be impossible for this sensor. The Kinetic sensor have a random error std of 17mm. I would expect the table to appear bumpy and the wire to appear as part of the table.
@@robotfpv Yes, seems you're right. From this paper www.ncbi.nlm.nih.gov/pmc/articles/PMC7827245/ I was hoping the NFOV std at close range would be suitable with additional processing techniques. Anyway, I will borrow the sensor and try a few different settings and techniques :)
@@robotfpvI’m interested in the fact Microsoft decided to revive the Xbox kinect for LiDAR development kits. I think devs would be more inclined to use LG Innotek hardware. But still cool. I love Microsoft. I just wish they would actually commit to a project for once. They never seem to last more than five years.
Really loved your content. I myself have recently started playing around with kinect, but facing a lot of difficulty cause of lack of documentation. It would be really helpful if you could guide me through some of the initial journey to build custom applications using kinect.
@@robotfpv I am trying to learn to implement Kinect from the very basics to build custom applications. I am working with C# with WPF apps, but open to other tech stacks as needed. Just would like some starting point.
Unfortunately C# is a bit a of a knowledge gap for me. Depending on what kind of application you want to build. You might be interested in depthkit for volumetric capture, or using kinect to build body tracking
Do you mean the OAK-D Lite? That camera also does depth perception by solving disparity. Is depth perception not the most "useful thing" about the Kinect DK?
Excellent and informative video! I work with signed languages, so the languages of the Deaf community, and I was wondering if I could ask your advice - given your expertise on motion capture. Signed languages use movement, hand shape, space, arm, hand, face, eyes, mouth etc - and even body motion. Our goal is to create an online dictionary of signs… but we also want to do more in depth linguistic analysis down the road. Other sign researchers use Azure Kinect DK, which is how I found your video. If you aren’t familiar with a signed language this vid will give you an idea of the dynamic nature of what we’d like to capture as effective as possible. ruclips.net/video/04tOoXCqyEM/видео.htmlsi=pieDljlHV0VNl7nA Given your expertise - what kind of camera or device would you recommend? We will be using a studio set up with a backdrop for most recording, other times, with focus sessions at Deaf events, it may be more uneven lighting - but with a backdrop. We’d like to capture as much fine-grained data as possible. Any suggestions would be greatly appreciated. Thanks!
Incredibly in-depth dive into the limitations and range of what the sensor can and can't do. Thank you so much for this!
Glad you enjoyed it!
Great comparison video. Please do upload more videos.
Just curious, when a depth sensor is in use as a camera, there are these black balls the fly through the frame, different sizes and speeds. In your example footage you see a few of these. They don't appear to be typical noise, so what are these black balls?
Great video. Subscribed. With this quality of content, you deserve to grow the channel.
Hello, what sensor would you recommend for 3d scanning boats. I have a friend that uses the Intel realsence, but I wonder if there is a better alternative out there in its price range.
Thank you! I can't decide between the Femto Mega and the Astra 2!
Hi! I just bougth an AzureKinect DK but i can't save image and record videos? How did you get it?
Thanks for the great video. How would it perform if it takes a point cloud of an object high contrast? Like highly reflective ropes wrapping around a black object.
Great video please do some more where you use some applications using these sensors
Hey, thank you for this, I was thinking about getting the Azure exactly for depth maps, stationary camera but moving circus people. Now I'm not so certain. Can you recommend something better in a similar price range? Or am I looking for more expensive equipment?
Thank you for the overview, really intelligent and helpful insights!
🙏 Thank you for your kind words
Hi, thanks for the review, this was very helpful! I would love to watch a comparison of all the sensors you mentioned, the robotics club at my school needs to purchase a good depth sensor but we aren't sure which one to buy. Thanks again, loved this vid!
Great suggestion! Is there another sensor you are thinking about?
Would love to hear your thoughts on how well these time sync to create one large depth sensor that performs body tracking in a large room
Really interesting video! Especially the part on reconstruction artifices. I was hoping to develop an application that can segment coloured cables in a scene with the RGB sensor. Then extract depth information to construct a spline for each cable. Do you think this parallax effect would cause many issues?
Glad it was helpful!
This depends heavily on the context of the application. Bare in mind that the sensors (IR and RGB) are only horizontally displaced. So if you had the cables placed horizontally and separately vertically, you might not see this parallax effect at all. But if you place the cables vertically, this parallax effect will definitely impact the depth sensing, especially if the cables are thin.
@@robotfpv Thanks! Yea, unfortunately the cables are 3 to 1mm in diameter. However, they will definitely be placed horizontally and separated vertically. They are just placed on a table so I can mount the Kinect anywhere and control lighting. I am wondering if capturing depth information on a 1mm diameter cable is unrealistic with this sensor.
@@Fergus858 1mm will be impossible for this sensor. The Kinetic sensor have a random error std of 17mm. I would expect the table to appear bumpy and the wire to appear as part of the table.
@@Fergus858 docs.microsoft.com/en-us/azure/kinect-dk/hardware-specification
@@robotfpv Yes, seems you're right. From this paper www.ncbi.nlm.nih.gov/pmc/articles/PMC7827245/ I was hoping the NFOV std at close range would be suitable with additional processing techniques. Anyway, I will borrow the sensor and try a few different settings and techniques :)
Hello, can I use one or more Azure Kinect to scan objects for 3D printing? Or would an Xbox Kinect 2 be better for that?
Thanks!
What are these things are used for? Motion capture... anything else ?
Robotics is one that comes to mind, especially for localisation, mapping, and collision avoidance.
@@robotfpv what mapping is in this context? The 3D scanning of the whole area ?
@@PeterMacLeod88 In the simplest form yes. Construct a 3D space of the environment so a robot can reasonably navigate and avoid collisions.
@@robotfpvI’m interested in the fact Microsoft decided to revive the Xbox kinect for LiDAR development kits. I think devs would be more inclined to use LG Innotek hardware. But still cool. I love Microsoft. I just wish they would actually commit to a project for once. They never seem to last more than five years.
Excellent video, thanks! Liked and subscribed
thanks man, appreciated!
More reviews please!
Wow, very high quality video. It seems like this system is getting no love from Microsoft. It seems like a huge step up
Deberíon sacar este artículo para las nuevas generaciones de consolas para jugar Just Dance.
With all the Facebook research they should be a player in this field.
Really loved your content. I myself have recently started playing around with kinect, but facing a lot of difficulty cause of lack of documentation. It would be really helpful if you could guide me through some of the initial journey to build custom applications using kinect.
Would be happy to help! What kind of application are you building? an what's a tech stack you are comfortable with?
@@robotfpv I am trying to learn to implement Kinect from the very basics to build custom applications. I am working with C# with WPF apps, but open to other tech stacks as needed. Just would like some starting point.
Unfortunately C# is a bit a of a knowledge gap for me. Depending on what kind of application you want to build. You might be interested in depthkit for volumetric capture, or using kinect to build body tracking
@@robotfpv Sure I would love to learn more about how to leverage DepthKit for volume capture and body tracking, if you could guide me through.
5:51wait, how?
They just revived realsense there are new models up right now of the 400x series.
if you know how to program, then you know how to use all sensors to limit the stuff you don't want.
Oak-D Light is what that little thing is and NO it doesn’t do what Kinect DK does, he failed to show the most useful things about the DK. 😮😂
Do you mean the OAK-D Lite? That camera also does depth perception by solving disparity. Is depth perception not the most "useful thing" about the Kinect DK?
Excellent and informative video!
I work with signed languages, so the languages of the Deaf community, and I was wondering if I could ask your advice - given your expertise on motion capture.
Signed languages use movement, hand shape, space, arm, hand, face, eyes, mouth etc - and even body motion.
Our goal is to create an online dictionary of signs… but we also want to do more in depth linguistic analysis down the road.
Other sign researchers use Azure Kinect DK, which is how I found your video.
If you aren’t familiar with a signed language this vid will give you an idea of the dynamic nature of what we’d like to capture as effective as possible.
ruclips.net/video/04tOoXCqyEM/видео.htmlsi=pieDljlHV0VNl7nA
Given your expertise - what kind of camera or device would you recommend? We will be using a studio set up with a backdrop for most recording, other times, with focus sessions at Deaf events, it may be more uneven lighting - but with a backdrop. We’d like to capture as much fine-grained data as possible.
Any suggestions would be greatly appreciated.
Thanks!