Our SDK is agnostic on how the images are acquired, but it requires the input to be in the cv::Mat format. More information on this format can be found HERE
The SDK comes with an example which uses the OpenCV highgui library to capture from video, images, or video streams. Compatibility with the IP/integrated cameras and video formats mainly depends on the device and on the version of OpenCV which is used for capturing. You can find the documentation for this handy library HERE
The SDK also provide functions to convert from most common image formats to the required one.
After an image is processed by the SDK, output is made available to the runtime in form of numbers.
Depending on the envisioned application, the programmer can use these numbers for his purposes.
For instance, you can write them down into a file, with your own specified format, save them in an online database, or activate / control specific functions.
More information about the output formats of our SDK can be found in the documentation and in the provided header files.