A Fast-Compressive Tracking Integrated with Differential Evolution to Optimize Object Tracking Performance

It is difficult to develop a desired efficient tracking algorithm and model for object tracking because each tracker has its dominant attributes and weakness. This work chooses fast-compressive tracking for its rapid real-time tracking with the element of compressed sensing theory. In addition, an evolutionary computation method, differential evolution (DE), is introduced to search for the optimal parameters by iteratively solving and improving a solution with a proper objective function. Hence, the proposed fast-compressive tracking integrated with differential evolution (FCTDE) has improved the object tracking performance. Furthermore, the integration has resulted in a reduction in the computational complexity of the tracking procedure. The experimental results are recorded on the standard Online Tracking Benchmark (OTB100) data sets to represent the robustness of the proposed tracker. As a result, the proposed improvement algorithm's average overlapping ratio and center location error (CLE) improved to 0.51 and 28.87. Thus, the proposed model enhances the ability to tackle occlusion and out-of-plane rotation problems.