To address the target coverage path planning (TCPP) in the background of inventory count, a two-layer optimization framework integrating the A* algorithm and improved particle swarm optimization (A*-IPSO) based on clustering is developed. Firstly, we propose a novel inventory count mathematical model, considering identification range of radio frequency identification (RFID) and obstacle avoidance constraints. Secondly, the path points generated by the A* algorithm are integrated into the framework, guaranteeing global optimality while effectively avoiding obstacles. Finally, to overcome the limitation of traditional PSO algorithms prone to local minimum, we propose an improved PSO (IPSO) with a nearest neighbor search (NNS) strategy to enhance solution quality and convergence efficiency in path planning. Experimental results demonstrate that the proposed method shortens the path length, improves computational efficiency, ensures obstacle avoidance, and outperforms traditional approach.