Video is actually a sequence of frames played in a continuous motion. Videos play a major role in day-to-day life. YouTube and Social medias mainly depend on the video data. The video should be short and precise, so that the people will show interest in watching them fully. A shot is a scene change from one to another. Specifically, a continuous recording done using a single camera is said to be a shot. In order to break a long video into chunks, it is necessary to find out the shot boundaries. By finding out the shot boundaries, it is easy to interpret a scene with the help of start and end frame of a particular shot. The scene change is broadly divided into two types: 1) Abrupt change and 2) Gradual (wipe, fade-in, fade-out and dissolve) change. In this work, a sliding window method is used in both machine learning and deep learning techniques to find out the shots. The performance measures such as, precision, recall and F1-score is used on the results obtained from standard BBC planet earth dataset.