In this paper, the reconnaissance mission, which is one of the critical aspects of warfare pertaining to air combat, has been studied. Amidst the enemy’s tight surveillance system, a swarm of heterogeneous UAVs needs to be deployed to perform the reconnaissance. For higher accuracy, a UAV must be as close as possible to the potential target location. However, this may lead to its exposure and destruction eventually. So, it is not possible for UAVs to inspect all the potential targets within a limited time. The main objective is to inspect as many targets as possible from a given set of targets without compromising the safety of UAVs. We propose a modified version of the Particle Swarm Optimization (PSO) algorithm to reduce the mean computational time and safeguard the UAVs for reconnaissance. The algorithm is implemented and deployed across three different mission environments with varying numbers of UAVs and targets. The performance of this algorithm is compared with that of two popular SI algorithms, namely, Ant Colony Optimization (ACO) and PSO. Simulation results establish its effectiveness for safeguarding the UAVs. Compared to ACO and PSO, the proposed algorithm yields a reduction of more than 45 \(\%\) in the mean computational time.