Image acquistion setup with strobe and usb 3.0 camera

[kevin512]

I am currently building a setup to acquire images of fast-flowing liquid using a strobe light and USB 3.0 Camera. I am a beginner with the LabVIEW and image acquisition setup.

I have selected a camera that is USB 3.0 vision compliant. The camera has a specification of 2 MP and a global shutter. I have not purchased it yet.

Camera specification: Basler a2A1920-160umBAS, 1/2.3 in. format, C-Mount, 1920 x 1200, 160 fps, Monochrome, CMOS Global Sh...

I would like to ask the community how I should approach building the acquisition hardware and software setup. I need to synchronize the strobe and camera so that the camera shutter is open for a longer duration than the strobe duration. The strobe duration is 0.5 microseconds. I plan to capture images for 1Hz for starters then increase the speed as needed.

Hardware in hand: GenRad1538-A strobe, NI9402 with CDAQ module.

Planning to purchase: Camera, USB 3.0 cable (I am not sure if a separate trigger cable is required)

Appreciate it if you could share any important blogs and forum discussions in mind. Thank you for your time and suggestions.

[BruceAmmons]

It looks like you are planning on purchasing the hardware from Graftek (good choice).  Have you talked to them about this?  They are very good at helping you figure out what hardware you need and how you should set things up.

You will need a trigger cable.  You could send a trigger signal to the camera, which would start the exposure and could send a signal out that indicates the exposure is active.  This could be used to trigger your strobe.  If you set the exposure up for one microsecond, there will be plenty of time for the strobe to do its thing.  It also guarantees the exposure is active before the strobe goes off.  You might be able to set the camera to a fixed frame rate and eliminate the need for an input trigger to start each exposure.

Bruce

[1100fps]

Half a microsecond is pretty darn quick, you may find yourself light limited. Binning the camera at expensive resolution may help.

In the past I have used strobes from Advanced Illumination that over volt, the LEDs for short exposure to get more power out per shot.

That’s fine for low frame rates, but as you frame rates increases, the LEDs will start getting hot and failing, for that reason, the electronics will limit the amount of power you couldn’t put out as your frame rates go up using very equipment. 

The larger your field of view, the more energy you will need. You may end up requiring multiple strobes, or single strobe controller with multiple heads, 

Bruce is correct that setting your camera exposure time to one or 2 µs will give you plenty of time for the strobe to go off. The camera can be set to bring a line high when the shutter is open, which will trigger the strobe.

 with your camera shutter speed, set that fast you don’t have any issues with ambient light in the exposure, although you may choose to turn the room lights off, just so you can see what the strobe is lighting up.

You have a fun project ahead of you, feel free to reach out if I can be of assistance

[AmitShachaf]

Hi,

I was looking into using strobe light with a frame grabber. But you can do the same thing with USB 3.0 camera.

Difference is that in the case of Frame Grabber you program the frame grabber.  The Frame Grabber is programed to trigger the camera first and then to trigger the strobe light.

If you are using USB 3.0 camera then the signal to the strobe light goes from the camera.

The timing of light of the strobe is the actual exposure control. Not the camera shatter.

You need to program the camera in that case and not the Frame Grabber. 

So, you need to check that LabVIEW driver for that camera gives you the control you need.

If you need more light, then what LED gives you I was looking into Hamamatsu Xenon strobes.

Good luck with fun project. 

[kevin512]

Thank you, Bruce, for your response. Now I am confident that the purchase would be good enough. I have not talked to Graftek yet but will contact them soon.

I planned to run the camera at fixed frame rates (low speed) and adjust the strobe flash duration. I took some preliminary images (several tries) with DSLR and strobe just to test the concept. Just need to automate it with the small camera now.

[kevin512]

Thank you for these wonderful suggestions.

I am using Xenon-flash lamp strobes instead of LED. With my preliminary images, I am quite happy with it although it is old technology. 

GenRad / General Radio 1538-A Stroboscope -IET Labs, Inc

I agree with low flash duration, the system may turn out to be light-limited. I can increase the flash duration to 3 us (again low frame rates). I am currently exploring some other options as well such as larger pixel size USB Camera 3.0 | 2.3 Megapixel IMX249 Mono | MER2-231-41U3M – GeT Cameras Inc (industrial-cameras.co...

Camera will be fitted inside a closed stainless-steel vessel. Hence ambient light would not be interference.

I will look into optimizing the camera size and FOV....need to explore lens and sensor combination. Thank you for pointing this out.

[kevin512]

Thank you for your help. I am not sure if I understand your response correctly. Kindly correct me if I am wrong.

I am not planning to use a frame grabber since it is an extra hardware cost (camera+ frame grabber). I prefer USB 3.0 since it can be easily connected to a lab PC.

I agree on the shutter part. Timing is controlled by the strobe, not the camera. In terms of synchronization, I want to trigger the camera first...then the strobe...then the camera shutter...then the next frame.

Please suggest if it is not possible to do so with USB 3.0. Are there any challenges in setting it up in Labview?

[AmitShachaf]

Sorry if I wasn't clear. You can do what you want with USB 3.0 camera and Strobe Light should work fine.

Only thing you need to check that IMAQdx has the control of the camera in the way you need. 

Frame Grabber could be easier if you buy the Frame Grabber from NI. 

Both options should work just fine. 

[1100fps]

I have not used the GenRad strobe since the '70's  Good old Strobotac.  That put a smile on my face.

Depending on number of pixels and desired frame rate, USB3 should work just fine. For long data runs, then look into a fiber extension, or Ethernet (GigE).  

I used Basler, Imperx, and Prosilica for strobographic work.  For my work, there was usually a trigger, but it could be an internal clock on the PC.  The camera needed to have I/O outside of the USB3 Data for image.  Generally, machine vision cameras have a little round Hirose connector out the back that have the I/0.  The fun bit is drilling down the feature list of the camera and setting up the parameters for the pins.

The trigger would send a signal to the camera to open the shutter, which would take a microsecond or so.  Once the shutter was opened on the camera, it would send a signal to the strobe to make with the light.

The GET camera has the needed I/O.  I could not tell from a quick peek at the web page if the camera could be binned. 

I strongly recommend a camera that conforms to A3 Vision Standards (Vision Standards | Automate).  That makes LabVIEW integration much easier. USB3 Vision is the standard for USB cameras.  It was not clear if that was a feature with the GET camera.

In the attached block diagram, there is a dotted line from the computer to the strobe controller.  The GenRad was invented before computers, so there is no programming option, but controllers like the ones from Advanced Illumination have a number of options that can be set, including illumination time.  Once set, you don't need that connection.

While it's probably not an issue, it's worth noting that a discharge lamp generates light by turning all the energy in a capacitor into light. The exposure time will not be a square wave, but a wave shape with a fast rise and a tail. LED strobes are typically controlled by FET solid state switches that generate an almost square wave output. 

Good luck, if you can, post results.  🙂

Attached a slide from a presentation I gave at NI week a LONG time ago.   The original talk was taking pictures of rotating equipment with tip velocities over mach1, but the principle remains the same.  Either your trigger is a sensor or a clock.