Setting up Basler Development Environment
Developing applications that make use of Basler cameras means to develop C++ programs which include and make use of the Pylon SDK.
The 3D Lab currently has 10 Basler dart daA1600-60uc cameras. Each camera connects via USB 3.0 and accepts an optional synchronization signal.
The cameras can be controlled directly through Basler's PylonViewer app, which works on Windows 7+ only... And, when connected to the Jetson TK1s, they are controled via our developed control app.
This page, however, is about the development of applications that use the cameras.
Note that the development platform does not need to attach to a camera; however you will need the pylonSDK on the development platform (for verifying the build process) and some computer with a Basler camera connected, to run the app on.
Prerequisites
- A Basler camera connected to a Linux computer.
- A Linux-based development computer.
- A C++ IDE (Netbeans, Nsight etc...)
- Access to the 3DLab SVN Repository.
Setting up
Download the Pylon SDK on your local computer. WE hav a copy of the SDK, version 5.0.12, for x86-64 architectures (64-bit desktop processors), in the repository ( 3DLab/code/pylonSDK ). Download and extract the archive, and follow Basler's installation instructions.
Alternatively, download the Pylon SDK from Basler's website. The Jetsons use Pylon version 5.0.5, installed in the default /opt/pylon5 location. To make development easier, the same setup ought to be replicated on your device.
In your development IDE, make sure to include the necessary library paths when including Pylon in your projects. The project 3DLab/code/camera-test/GrabImage is a nice, simple pylon-enabled Netbeans application that demonstrates the basic Pylon usage and the necessary include/dependency structure.
Testing
Same as any other 3DLab camera-system project; commit your project to the repository, check out the project via Jetson's commandline, build on Jetson, launch on Jetson.
Since Jetsons are running in headless mode 99% of the time, don't plan yoru applications around showing a test image on the Jetsons' display. Instead, write images/data to binary files, or log debug information to cout/cerr/logfiles.
commit: $Id: howto-setup-basler-devenv.md 817 2018-10-12 12:34:32Z elidim $