A method includes receiving an indication that a video mode is invoked, where the receiving causes the following operations to be automatically performed: obtaining a first frame and a second frame at a first frame rate; defining a first ROI in the first frame based on a user's selection; defining a second ROI in the second frame based on the first ROI; determining a first camera motion flow; determining a first ROI motion flow between the first ROI and a corresponding portion of the second ROI; determining a second frame rate based on a first comparative value determined using the first ROI motion flow and the first camera motion flow; and capturing a third frame at the second frame rate by an image sensor unit, or playing back a fourth frame at the second frame rate on a touch screen unit.