A system for serving an advertisement in a networked environment receives data that defines a user selection of an image. The system also selects an advertisement associated with the user selection and communicates the selected advertisement to the user. The data that defines the user selection includes an x selection coordinate, a y selection coordinate, a width, and a height that defines a region of an image. The user selection defines also defines a scene of a video. The system also includes circuitry and logic configured to extract text and to detect objects from an image region defined by the user selection and to select an advertisement associated with any extracted text and/or detected object.