Low Cost Lidar crowdfunding suggestions

[Pippy]

Would be willing to send a link to the filter you are using?

We searched for a long time to find a camera module that fit our needs, and it is color, so we can't really change that....

http://www.ebay.co.uk/itm/9x9x1-0mm-650nm-laser-High-transmittance-Filter-Against-400-1100nm-/261057674413?pt=LH_DefaultDomain_0&hash=item3cc83ffcad

If you put the filter on a colour camera you'll no longer see anything but 650nM - laser red colour. No blue or green will get through.

[Maeth]

Cool, thanks a lot!

We are already set on our sensor, but maybe others would want to know of what sensor you are using?

[Maeth]

We want to maintain the color stills and video feature for the user, so we can't use the filter.

However, we set our exposure similar to you, such that the laser is consistently the brightest thing in the image.
We also do limiting on size, so that the sun or a bright window are not counted as part of our laser dot.

[Pippy]

Well the laser dot only falls onto a narrow strip of pixels across the sensor so a lot can be done to filter out the background.

The monochrome sensor I'm using is the Micron MT9V022. When setting it's ROI (Region Of Interest) to say 32 pixels high right across the sensor you get up to 250 to 300 fps readout, which is very nice.

If you toggle the laser on/off every other frame (one frame with laser on, next frame with laser off) you have no trouble at all in finding the laser spot and rejecting EVERYTHING else (especially at 250fps), the spot becomes obvious even in high brightness scenes where you might get other bright spots from something or other that look just like the laser spot. Makes things far less error prone when it comes to your algorithms in your firmware.

And when doing sub-pixel center of gravity on the laser spot it has a high distance accuracy.

The reasons I chose a monochrome sensor over a colour one is because the laser spot can be filtered (with the glass filter) and because you don't have the bayer pixel problem you get with colour sensors. Because of the bayer matrix on a colour sensor you only have a 1/4 of the all the pixels that actually detected the 650nM red spot, which of cause reduces your image/distance resolution. With a monochrome sensor all pixels around the spot contribute to the distance accuracy.

[Maeth]

Sounds like you are doing a similar project to us! Awesome!

We did the laser/no laser trick early on, but this cuts your refresh rate in half so it is not very desirable.

You got it!

[Maeth]

Yep, we are currently looking into using a line sensor, so that multiple points above and below the plane are registered simultaneously.

Also, as we have an IMU, you will be able to reconstruct the scene in 3D on a computer!