Path planning is key to ensuring that unmanned aerial vehicle (UAV) can complete its missions safely. It is necessary to plan a path between the starting and end points that allows UAVs to avoid no-fly zones and threat sources while meeting their own dynamic constraints. To address the issues of slow convergence speed, susceptibility to local optima, and insufficient adaptability of standard artificial bee colony (ABC) algorithms in complex environments, an improved artificial bee colony (IABC) algorithm is proposed in this paper. Firstly, the population is initialized using chaotic mapping to ensure that initial solutions are more evenly distributed throughout the entire solution space, ensuring that they cover the area where the optimal solution is located. Then, to accelerate the convergence of the algorithm, the search strategy at the employed bee stage is enhanced to maintain a strong global search capability towards the optimal solution. Concurrently, the search strategy of the onlooker bee stage is also improved, endowing the improved algorithm with a stronger ability to explore local information. Finally, path points are interpolated and smoothed. This 0paper compares the proposed IABC algorithm with the basic ABC algorithm and the global best ABC (GABC) algorithm. The experiments results demonstrate that the proposed algorithm has good effectiveness and feasibility in solving the UAV path planning problem.