Blogs

Playing with yolo filter !

Up till now I have been porting scans over the Serial wires and manually presenting the data.....

Now the Rig has its own Real-Time Debug TFT Display. (you might see it below the right wheel)

Above is just a 180° scan of this below :-

Wooden planks set at 90° and two DVD cases thrown in for measure.

When Richard Ryerson (RobotGuy6)  takes two upper InMoov body to build a dog.

Cute name: FrankenDog according to Richard.

Richard is running EZ. I was lucky to stumble on this video, because there is no mention of InMoov on YouTube.

Okay now lets see if it opens a door and dance like the Boston Dynamics dogs!

:)

I placed the Sensor rig into a cardboard box just to get a feeling for the data.

This is two passes of the probe, so you can see the slight variance in data.

Overlaying the raw data for 180°.... this was with GIMP so its just approximate

all looks kinda good except Hmmmm.... rounded corners cannot be explained...

another scan with DVD's :-

Its scanning, the stealthy Dual Time of Flight wiring was a soldering pig to execute (or should I say "Exterminate").

Data is flowing.... at momo a bit strange 200 Steps = 360° (until microstep instigated giving theoretical 1/16 steps ermm if my maths is correct 3200 Steps for 360° however at the price of speed)

Two is better than one...... (or 4 or 8 etcetcetc).

Following successful experiments with 1 single Time of Flight distance sensor mounted to Stepper, watching the scan process its seems to lend its self to setting the scan to just 180° and mount two sensors back to back to extract the whole 360°...... (or 90° scan with 4 sensors, hope you see where I am going with this :-)

Alex , they beat you to the dancing bot :)