Machine Vision Camera SDK (MvCameraSDK) provides multiple general APIs supporting protocols such as GigE vision, USB3 vision, Camera Link, CoaXPress, and XoFLink. You can control the parameter configuration and image processing logic of different cameras to quickly set up vision applications that meet your needs.

The basic features of MvCameraSDK consist of parameter configuration, image acquisition, image processing, and event.

• Parameter configuration: Supports getting/setting all parameters of device, including width, height, and exposure time.
• Image acquisition: The SDK supports receiving images from the machine vision camera, and provides active search and callback modes for upper-level software.
• Image processing: Supports rendering image, converting format, saving image, recording, etc.
• Event feature: Supports device events and abnormal status uploading.

• Linux Operating System (32/64-bit): Ubuntu 16.04 to 24.04 (32/64-bit), Redhat7 (64-bit), Centos7 (32/64-bit), gcc/g++ 4.8.5 or later.

  Note

  The frame grabbers does not support 32-bit system, For frame grabbers, only 64-bit Ubuntu 16.04 to 24.04, and gcc/g++ 4.8.5 or later are supported.

• Python Version: 2.X, 3.X

The client is installed in /opt/(software name)/ by default. After installation, you can find the dependent files of the SDK in the Development folder. The content is as follows:

• Bin (complied deliverable)
• Documentations (documents for secondary development, such as SDK and user manual)
• Includes (header files)
• Lib (dynamic link libraries)
• Samples (sample programs in different programming languages)

  Note

  The Python header files are located in the .\Python\MvImport directory under this folder.

Note

Files under development directories of different client version might vary. The actual circumstance shall prevail.

Machine Vision Camera SDK Linux Version includes driver and logserver (log service). You can acquire them in the installation path.

• The sample programs provided are only for reference. Improper codes may fail to implement the expected camera feature or even cause damage to the application. Therefore, you should adjust the sample program according to your specified needs and test it before use.
• The APIs provided are only for reference, and the actual supported APIs for cameras shall prevail.