This is a framework for video analysis and structurization, written in C++ with few dependencies and easy to get started. It works like a pipeline where each node is independent, allowing for the creation of various types of video analysis pipelines. It is suitable for scenarios such as video structurization, image search, facial recognition, and behavioral analysis in security fields (e.g., traffic incident detection).