Dynamic arbitrary shape boundary detection anc tracking using Hungarian Kalman filter

The paper explains about tracking a dynamically changing arbitrary shaped structure, important for determining their boundary and position estimation for advanced warning. The proposed algorithm uses active contour detection algorithm to construct the effective contour in each frame. Then Hungarian based Kalman filter is used to achieve best accuracy in estimation of future positions using multi-point tracking with cost minimization assignment problem. An arbitrary shape shadow is constructed on photo-voltaic cell as an experimental set up and dynamic arbitrary shaped shadow is generated. Results of the algorithm shows 98% area accuracy of estimated and original shadow positioning.