Video transfer through a CAN Bus

CAN is designed as a real-time instrumentation and control bus, not file transfer or data streaming. It's packet size is small making it unsuited for either role in general (it has a large per packet overhead). You can do file transfer, and some do, particularly for reprogramming a part or dealing with configuration of fault data. Your top end data rate is probably a few hundred kilobits per second.

It has been tried. Take a looks at

www.coe.montana.edu/.../canbus_video

I have no idea how the project fared.

Robert

Hello Robert

Interesting. Does not mention the speed of operation (CAN perform at higher rates when at short distances)

Regards
Amit

Amit,

They do mention a CAN Bus rate of 1Gbit/s in one place. That's 3 orders of magnitude too high for even the raw baud rate. I suspect/hope that's a typo but it does suggest they are expecting a maximum speed bus. Given the application is on the end of a sprayer boom I would expect a rather lower rate. Also no suggestion of the bandwidth required by their choice of camera.

It's a student project. Still learning about what questions to ask I suspect.

Anyone seriously interested might be able to find out how the project went if they contacted the faculty.

Robert

Hello Robert

Yes. Also the camera is a 115.2Kbps with JPEG compression. At this baud rate and considering the size of image the maths does not add up to 1Gbps

Regards,
Amit

May I register as, "Doubtful poster #3?"    Commercial products exist in abundance - and usually employ the most time-tested/proven communication medium.   (which I do not believe is CAN - due to packet size - and other limitations.)

The fact that something "can" be done does not mean that it "should" be done...  (pun intended)

Robert Adsett said:

It's a student project. Still learning about what questions to ask I suspect.

Anyone seriously interested might be able to find out how the project went if they contacted the faculty.

 HI Robert, as all is posted here seems not have a future, primary it is dated 06 and not least sponsor site is broken. Maybe this idea get them out of business.

@CB1, in wich sense are you doubtful?

Roberto Romano said:

@CB1, in wich sense are you doubtful?

Only in the sense of, "Best tool for the job" - which (as stated) due to the "limited packet size" - appears "not" the best/brightest means to transfer (reasonably sized) video data rates...

It appears to be an undergraduate project, so neither of those disconcert me. In such projects the web site is typically a side concern, often never updated.
It's not unlikely the project turned into more work than the time allotted or they ran into the problems we've alluded to. In either case if the project was completed then the final report may be available.

Robert

IMHO such a project is technically feasable, if one accepts two premises:

• realtime capability is impossible for any decent resolution and frame rate
• the protocol overhead is rather large, 50% would be a first ballpark figure

That would leave not much incentive for a project with any commercial  background.

Poster's title, "Video transfer through a CAN Bus."

Appears we're in agreement - simply because a method "can" be employed does not insure that it "should" be employed.

Video transfer through a CAN Bus is an improper mating of source to tool - bad idea...