The exploration of planetary surfaces is predominately unmanned, calling for a landing vehicle and an autonomous and/or teleoperated rover. Artificial intelligence and machine learning techniques can be leveraged for better mission planning. This paper describes the coordinated use of both global navigation and metaheuristic optimization algorithms to plan the safe, efficient missions. The aim is to determine the least-cost combination of a safe landing zone (LZ) and global path plan, where avoiding terrain hazards for the lander and rover minimizes cost. Computer vision methods were used to identify surface craters, mounds, and rocks as obstacles. Multiple search methods were investigated for the rover global path plan. Several combinatorial optimization algorithms were implemented to select the shortest distance path as the preferred mission plan. Simulations were run for a sample Google Lunar X Prize mission. The result of this study is an optimization scheme that path plans with the A* search method, and uses simulated annealing to select ideal LZ-path- goal combination for the mission. Simulation results show the methods are effective in minimizing the risk of hazards and increasing efficiency. This paper is specific to a lunar mission, but the resulting architecture may be applied to a large variety of planetary missions and rovers.