The camera does v4l2
01-06-2021, 10:55 AM
(This post was last modified: 01-10-2021, 11:33 AM by ab1jx.
Edit Reason: typo
)
There, try this:
#!/bin/bash
adate=`date +"%Y-%m-%d_%H-%M"`
uvccapture -d/dev/video2 -v -x1600 -y1200 -o$adate.jpg
Needs uvccapture, it's in the debs. Not sure about exposure control, the sun's coming in the window and it looks overexposed. Why it's not video4 I'm not sure, that didn't work so I kept stepping down one at a time until one worked. The Raspberry Pi cameras send on different "ports", not sure if that's like a TCP/IP port or not. video4 may be for actual video.
My actual first picture:
There's a guy I stumbled across, I think in Australia, that does a lot with v4l2, virtual reality, OpenGL. I think he works for Google now. He does object recognition more by shape than by looking at pixels. Uses those cool visors with tiny screens inside, wearable computers. So finding this does v4l is very significant. Focusing would be good, I have cell phone cameras that can focus.
This guy: http://www.tinmith.net/ Dr Wayne Piekarski
#!/bin/bash
adate=`date +"%Y-%m-%d_%H-%M"`
uvccapture -d/dev/video2 -v -x1600 -y1200 -o$adate.jpg
Needs uvccapture, it's in the debs. Not sure about exposure control, the sun's coming in the window and it looks overexposed. Why it's not video4 I'm not sure, that didn't work so I kept stepping down one at a time until one worked. The Raspberry Pi cameras send on different "ports", not sure if that's like a TCP/IP port or not. video4 may be for actual video.
My actual first picture:
There's a guy I stumbled across, I think in Australia, that does a lot with v4l2, virtual reality, OpenGL. I think he works for Google now. He does object recognition more by shape than by looking at pixels. Uses those cool visors with tiny screens inside, wearable computers. So finding this does v4l is very significant. Focusing would be good, I have cell phone cameras that can focus.
This guy: http://www.tinmith.net/ Dr Wayne Piekarski