harvesters

Why we included this project

Most vision projects that touch industrial cameras start with the same chore: talking to the hardware through GenICam. Harvesters handles that layer for you in Python, wrapping the GenTL producer interface so you load a transport library, find the connected cameras, and pull frames in a few lines of code, with the data coming back as NumPy arrays ready for your processing pipeline. Because the GenICam standard is shared across vendors, the same API works whether your cameras talk GigE, USB3, or another GenTL transport, which is handy when one application mixes brands. You also get the feature node map, so exposure, gain, and pixel format can be tuned from code rather than through vendor GUIs. For inspection, robotics, or research on real hardware, that keeps acquisition code portable and saves you from rewriting it when you swap cameras.