The growing use of multi-robots in various fields highlights their increasing importance. These robots can tackle a range of challenges, including the pursuit-evasion problem. When dealing with multi-robot systems for the pursuit-evasion problem, precise estimation, planning, and tracking are essential. Our approach offers a factor graph-based solution to this problem, focusing on scenarios with one evader and multiple pursuers. Through a series of experiments, both simulated and with real hardware, we demonstrate the effectiveness of our algorithm. It enables accurate estimation and tracking of the evader, allowing pursuers to efficiently capture them in the shortest time and distance compared to standard methods.