Fractional Order Derivatives as an Optimization Tool for Object Detection and Tracking Algorithms

Optimization techniques are nowadays one of the most important branches computer science due to limitations computing power availability especially in cases mobile platforms. Algorithms which can be used on such units have optimized for memory and power. In paper we focused optimization image analysis algorithms by limiting number necessary operations. The algorithm has been designed detect follow a specified marker, known a priori , is based correlation coefficient match between acquired template current image. taken from previous processed frame. operation Pearson’s coefficient, so whole mechanism therefore highly demanding terms performed mainly using Region Interest (ROI) exclude irrelevant parts Grünvald - Letnikov fractional order backward difference estimate position marker sequence images. This limits operations required maintain precision algorithm. Based object frames, mathematical tool able at relatively low cost with high accuracy probable incoming Here, explain workflow detection following algorithm, as well basis derivative tool. processor load connected was reduced over 35%.