There is provided a system and method for visual search in a video media player. There is provided a computing device for performing a visual search in video files comprising a display, a memory, a video media player application in the memory, and a processor configured to store into the memory search terms by allowing a user to visually select objects as search terms using a user interface on the display, initiate a search using the search terms, obtain search results comprising time offsets and play durations within the video files, each play duration containing at least one frame corresponding to the search terms, and display navigational controls in the user interface for presenting the search results, including selectable graphical thumbnail images or preview videos. A user can select objects via outlined, highlighted, or otherwise visually identified objects in a video frame, the selecting enabled by associated search metadata.