Object tracking is the cornerstone for all machine vision applications. The major challenge in object tracking applications has always been speed, accuracy and ability to process heavy data flow. By using field programmable gate arrays in implementing the algorithm, we can overcome the challenges primarily by using its ability of reconfigurable design and parallel architecture. The basic building blocks of this project are pre-processing, image splitting and detection followed by tracking and display. The pre-processing block uses scaling of image through bilinear interpolation so as to make it more applicable for computation, the image splitting and detection block do the splitting of image after compression, which is achieved by breaking the image into submatrices. Through consistent comparison and analysis of the stream of images with a default image, the object is tracked. © 2016 Pushpa Publishing House, Allahabad, India.