15.1.20. Wrap up

The IDE in one paragraph: scripts are edited in a professional editor that knows the camera’s API, run on the camera with one button, and observed through three live instruments – the frame buffer viewer for what the camera sees, the histogram for the numbers behind it, and the serial terminal for what the script says. Around that loop sit the maintenance tools that keep a camera’s firmware, filesystem, and ROMFS in order; the machine-vision tools that produce the artifacts scripts consume – threshold tuples, cleaned descriptors, printed tags, converted models, labelled datasets; and the power tools that look under the hood when performance matters.

Where to go from here:

• The examples menu is the standing answer to “how do I do X on the camera” – nearly every library feature has a runnable example.

• The library reference documents every module the completion popup offers.

• The openmv Python package drives a camera from host-side Python scripts – the IDE’s debug protocol without the IDE – for test rigs, automation, and custom desktop frontends.

• The production chapter picks up where the IDE’s deploy step leaves off: baking scripts into the firmware, shipping assets in ROMFS, and hardening a camera for the field.