
I have ideas about how to use stereoscopic cameras to run an automatic calibration routine, but haven't had time to roll out experiments. Your camera, printer, and slicer object rotation must all be very carefully calibrated and anchored. The 2D masking experiments I ran were low signal / high noise compared to 2D bounding boxes. It's possible to apply a segmentation model and "mask" expected real-world shape based on a mask extracted from the 3D source files.ģD segmentation proposals are expensive to calculate/reduce (imagine rendering a point cloud). I've tried a few experiments using 3D model source as a model input, and. Thank you! Your instincts are totally right. Hope this helps! It's a wild world out there. Moonraker communicates with "Klippy Host" via JSON RPC. Klipper also provides the ability to use your Raspberry Pi as a 2nd controller if your 3D printer's microcontroller doesn't have enough pin-outs to support your haĤ) Mainsail is a front-end webapp that talks to Moonraker, which is a back-end app that manages Files/Gcode/Print Job via HTTP, MQTT, and JSON RPC. I couldn't find a list of supported boards, but you can see all the usual suspects here (stm32, rp2040, etc). Klipper provides a JSON RPC API for communicating with your printer, without having to worry about serial baud rates or other low-level details.ģ) You must connect Klipper server (Pi) to the microcontroller running Klipper firmware.

1) You flash Klipper firmware to your 3D printer's microcontroller ("MCU" in Klipper docs).Ģ) You run Klipper's server software ("Klippy host") on a Raspberry Pi (or other single-board computer).
