MatrixVision is a video intelligence platform integrating two core capabilities of edge computing and deep learning. MatrixVision adapts to a variety of edge computing hardware, encapsulates core functions such as video stream decoding, image decoding and encoding, model conversion and migration, model inference, image detection, image classification as key function modules, and provides comprehensive and real time function interface for multiple scenarios.