Extract Motion from Picture Sequence

Optical flow is a concept for estimating motion of objects within a visual representation. Motion is represented as vectors at every pixel in a digital image sequence of two frames which are taken at time t and t + δt respectively. Based on the assumption of the observed brightness (intensity) of any object point is constant over time interval and the movement is small, optical flow equation (OFE) is derived and used in algorithm to calculate optical flow for a particular motion. In order to inject robustness in the algorithm, time-varying uniform illumination and calculation of large range of motion also have been taken into account while formulating the algorithm. The algorithm is used as reference to create a coding input for the MATLAB, which is the software used in this project. As a result, the software generates a separate view of static and moving objects. Further study of the output obtains more accurate data. The experimental process will be carried out using a video as an input to determine whether the project succeeds or need more modifications. The outcome of this project is beneficial for motion analyst to predict the parameter of moving objects, animations, statistics and also robotic eye.