Message from @ManAnimal

Discord ID: 614557162781278218


2019-08-23 20:25:29 UTC  

now need to convert in order to apply Open cv

2019-08-23 20:25:42 UTC  

Sometimes it worked, and sometimes just went horribly wrong

2019-08-23 20:26:03 UTC  

the idea is to take a video and use photogrammetry to create a 3d object

2019-08-23 20:26:08 UTC  

And with each revolution it would fuck up somewhere else, while fixing one spot

2019-08-23 20:26:12 UTC  

NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEELLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLL

2019-08-23 20:26:51 UTC  

FFFFFFFFFFFFFFFFFFFFFFFFFFFAAAAAAAAAAAAAAAAAAAARRRRRRRRRRRRRRRRRRRRRRRRRRAAAAAAAAAAAAAAAAGGGGGGGGGGGGGGGGEEEEEEEEEEEEEEEEEEEEE

2019-08-23 20:26:54 UTC  

cause 2D sections<plane> are no different that 4D frames <images>

2019-08-23 20:27:29 UTC  

3D imaging from 2D has some serious issues with shadows as it turns out

2019-08-23 20:27:48 UTC  

it does, but if you strip the back ground

2019-08-23 20:27:50 UTC  

A shadow makes it thing that the spot is deeper/shallower than it really is

2019-08-23 20:28:27 UTC  

it works a ton better if youu use OpenCV to detect countours and then strip everything but your target object

2019-08-23 20:28:41 UTC  

It works better when the camera is monochrome and the object is lit by an array of lasers

2019-08-23 20:28:43 UTC  

then run those frame through the SIFT algorithm

2019-08-23 20:28:55 UTC  

lidar is best

2019-08-23 20:29:16 UTC  

but you can strip the color out of the image with one command with opencv

2019-08-23 20:29:41 UTC  

A parrallel course to mine was making one such table thing that digitised phyiscal objects

2019-08-23 20:29:47 UTC  

lidar actually has a similar problem; coordinate systems

2019-08-23 20:30:03 UTC  

They got some serious backing from bosch

2019-08-23 20:30:29 UTC  

Like they not just used a laser array and a dome of monochrome cameras, oh no
that was too simple

2019-08-23 20:30:42 UTC  

Those lasers also measured distance each

2019-08-23 20:30:45 UTC  

yeah, I'm trying to integrate those two workflows so i can use photogrammetry info in cad and then laser data combined to photogrammetry

2019-08-23 20:31:28 UTC  

but everything uses different coordinate systems which aren't readily connected

2019-08-23 20:31:56 UTC  

In the end they abandoned the dome idea and went with a turntable that had an arch of cameras and lasers

2019-08-23 20:32:26 UTC  

Actually not an arch, more like almost a complete circle

2019-08-23 20:32:29 UTC  

a lidar array is the best way to go if you can afford the equipment

2019-08-23 20:32:33 UTC  

high accracy

2019-08-23 20:32:53 UTC  

it's just stitching such large datasets can be a bear

2019-08-23 20:33:33 UTC  

that's another area not connected; photo panaroma's and point cloud registration

2019-08-23 20:34:03 UTC  

One side contained lights and normal cameras (for colour data) and the other side had 3 horisontal laser lenses with an array of monochrome cameras and distance measuring inhole camera-chips

2019-08-23 20:34:08 UTC  

meme

2019-08-23 20:34:09 UTC  

MEME

2019-08-23 20:34:09 UTC  

there are some kick ass open source tools but they are all insulated from each other

2019-08-23 20:34:10 UTC  

meme

2019-08-23 20:34:12 UTC  

ME E M

2019-08-23 20:34:13 UTC  

mem

2019-08-23 20:34:14 UTC  

Rather smart reduction of component count

2019-08-23 20:34:17 UTC  

<:thonkvult:580055400376827905> don't you thikn?

2019-08-23 20:34:43 UTC  

my nigger

2019-08-23 20:35:10 UTC  

correct salute

2019-08-23 20:35:14 UTC  

Wrong audience