Camera Module Recommendation

mynet43 · 2012-02-12 19:55

I need a color camera module for a project I'm working on.

A resolution of 320x240 (or a little less) is OK.

It needs to have RGB 565 output.

It will be used to take single frame pictures and transmit them to a display device.

It needs to focus from about 2" to 4"+ and probably needs autofocus within this range.

I'll be using it on an electron microscope, to locate the target before extreme magnification.

Does anyone have a recommendation for a camera module that would work here?

Also, is there software available that would support this type of application?

Thank you for your help and support.

Jim

shimniok · 2012-02-13 11:29

This Toshiba camera from Sparkfun can be configured down to lower resolution. Not sure about 565 or focal range; you'll have to check the datasheet.

There's some info out on the web about how to use this camera if you dig around a little. If you have trouble finding stuff let me know. Despite one thread where a lot complain about not getting it to work, there are a few people who have and have the images to prove it. I have one of these and designed a breakout board based on sigalrm.blogspot.com's work, but using a dual-supply regulator to save board space. I have yet to actually try using the camera, though.

Leon · 2012-02-13 11:37

I've got a couple of those cameras, and keep meaning to do something with them. There is an Eagle PCB part for them, I've created one for the Pulsonix software I use which was based on it.

ericball · 2012-02-13 12:07

I was looking at the TCM8240MD, which is the 1300x1040 big brother to the 640x480 TCM8230MD and one item of note is both cameras provide continuous output. So you don't trigger a virtual shutter then poll the device to read the pixels. Instead the device generates a clock for the pixel data and your app has to accept it at that rate. (With the TCM8240MD I couldn't see any way to control the data rate, maybe there's more detail in the TCM8230MD datasheet.)

I don't know if a camera module with an onboard buffer exists. It would be much easier for my project idea (a variation on the panorama ball) if a microcontroller could trigger multiple cameras synchronously, then pull the image data asynchronously.

Leon · 2012-02-13 12:17

I don't think that the Propeller is fast enough for them. They are usually used with something like an ARM.

pedward · 2012-02-13 14:16

You feed the camera a clock, which determines the output rate, combined with the frame rate selector. Set it for a low frame rate and feed it the minimum clock value based on the datasheet's calculations. We discussed these a few weeks ago and I pointed to both of these cameras as possible solutions for Prop vision. The 1.3MP unit is more flexible than the .3MP unit.

Kye · 2012-02-13 16:16

Please wait for the CMUcam4 to come out in a couple of months. It will be able to do just what what.

Thanks,

mynet43 · 2012-02-13 17:37

@pedward,

Do you have code to make these cameras work?

The people on SparkFun didn't seem to be able to make them work, because of bad documentation and lack of support.

If you have them working, I'd sure like to see the code. It would be a big help right now.

@Kye,

Thanks for the tickler.

Will the CMUcam4 be able to focus down to 2" and will it have autofocus from 2" to 4"?

I need both of these for this project. I have a moveable target within this range.

I may not be able to wait two months. I have a customer now

Thanks everyone for the great responses.

Jim

Kye · 2012-02-13 18:28

No, I don't think it can focus on an object that close. Sorry,

It can take the picture though.

Good luck with other options.

The CMUcam4 looks like this. If you could mount something to it,,,

mynet43 · 2012-02-13 18:58

Thanks Kye,

That's a cool looking board. I may play with it for other projects.

Jim

mynet43 · 2012-02-13 19:14

@Michael Shimniok,

Do you have the code available from the person that got this camera working?

I would be very interested to see how they set the registers and use the commands.

Thanks,

Jim

Leon · 2012-02-13 19:22

This paper includes ARM code:

http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA499618

mynet43 · 2012-02-13 20:00

Hi Leon,

Thank you for the link to the paper.

A couple of notes after reading it:
1. They mentioned that there were around 95 registers affecting the picture that they were unable to use because of insufficient documentation.
2. The code seems to be somewhat incomplete, such as a number of 'include' files that are not shown.
3. The description makes it clear that the picture output is started by a single command. Once this is sent, simply monitoring the clock and H and V signals allow scan lines to be transferred smoothly.

This looks like it should be fairly easy to implement on the Propeller.

Since their goal was to capture gray scale images, it's hard to determine what the picture quality would be in full RGB color.

I really appreciate your sending this. It gives me a good idea how to interface with the camera. Since it doesn't have autofocus, I'll have to keep looking.

Thanks again,

Jim

shimniok · 2012-02-14 06:41

I ran across some more detailed Application Note kind of documentation on the camera not long ago. Click this: TCM8230MD_AplNote_E_040127.pdf

I don't have code but email the person at sigalrm.blogspots.com -- or search around for other implementations (I think one other person has recently posted a working board & software and such, can't remember where I saw it, though)

Michael

mynet43 · 2012-02-14 06:51

Michael,

Thank you for the link to that document.

It seems to have detailed descriptions of many of the registers, commands and timing.

Looks really good.

Thanks again,

Jim

mynet43 · 2012-02-14 07:04

Here's a link to a camera that might work.

http://www.sony.net/Products/SC-HP/cx_news/vol41/pdf/mcb770.pdf

It has autofocus down to 10 cm and seems to be able to output uncompressed data.

Let me know what you think. I plan to try to find out more about it.

Jim

Camera Module Recommendation

Comments