A video pre-processing front end is integrated into the SDK to provide de-interlacing, scaling and chroma decimation, as well as a suite of video analytic algorithms. The tight coupling of these video analytics with the CODEC core gives the Intelligent Encoder SDK the ability to dynamically adapt its behavior to the requirements of the video stream being processed.
This ability to dynamically adapt its parameters makes the Intelligent Encoder an ideal solution for applications such as video surveillance. When scenes contain little or no motion, the encoder can reduce frame rates or resolutions and reduce bandwidth or storage consumption. When motion is detected within the scene, the built-in analytics can direct the CODEC to increase frame rate, resolutions or compression settings to maintain consistent video quality. The API of the Intelligent Encoder allows privacy regions and regions of interest to be defined in order to focus where analytics are to be applied.
The Intelligent Encoder is capable of performing H.264 AVC, H.264 SVC, MPEG4 or MJPEG video compression with simultaneous G.711 audio compression. On S6100 and S6105 devices, the H.264 encoder supports 30fps streams with resolutions up to 1.3 megapixels. At standard definition, D1 resolutions, 75% of the processing capability of S6100 and S6105 devices remains available for third party applications or analytics. The rich API and modular nature of the Intelligent Encoder allow for easy integration of third party applications.
The Intelligent Encoder is also a powerful decode engine accepting input streams from both Stretch plug-ins and third party CODECs and producing decoded video for display. A built in scaling and captioning engine gives the Intelligent Encoder the ability to scale decoded or live video for tiled displays and to add user defined captions that can be alpha blended onto the video stream.
