The segmentation of multimedia data flows is a fundamental problem with many applications. Properly segmented flows can be better organised and reused. They provide access points that facilitate navigation and retrieval. As more and more multimedia data are created and made available, segmentation algorithms can serve the important function of helping to summarise this mass of material. There are several advantages of genetic algorithms over current segmentation methods, such as clustering . First, the genetic mechanism is independent of the prescribed evaluation function and can be adapted to support a variety of heuristic-based characterisations depending on gender, domain, type of user, and so on. Second, evolutionary algorithms are naturally suitable for doing incremental segmentation that can be applied to the transmission media . Third, it can support dynamically updated segmentation that adapts to usage patterns, such as adaptively increasing the likelihood that frequent access points will appear as segment boundaries. In this article we will focus on the video. This can produce raw video or video. Examples of video produced are news, movies and training videos. Examples of raw video are video meetings, surveillance records, and usable personal video cameras . The method described in this document can be applied to non-image data streams. The genetic segmentation algorithm remains the same; What is required are different fitness functions that take into account the appropriate characteristics of that medium and the software to process that medium.