Optical flow optimization using a parallel Genetic Algorithm
Introduction
Optical flow is the displacement field from one image into another. It shows the velocities from a 3D space projected into a 2D space and it is being used in many different areas like medical imaging, tracking and object segmentation, etc. This work presents a parallel genetic algo- rithm (GA) that estimates the optimal param- eters of an optical flow algorithm. The multi- channel gradient model (McGM), presented by Johnston et al.[3], is an optical flow algorithm that works well under different problems like static patterns, contrast invariance and differ- ent kinds of noise. This model depends on more than ten parameters which are being es- timated by the parallel GA.