Path planning has applied in many fields such as bioinformatics, CAD/CAM, layout, image processing and especially in robotics. Its target is to find out the best path with respect to some optimization criteria in order that an object can move along the path from a starting point to an ending point in two or higher dimension spaces avoiding a set of obstacles. Path planning mainly follows three approaches: search-based, sampling-based and combinatorial planning. In combinatorial planning, building a visibility graph is a main phase in the whole process and theoretically it takes Θ(nlogn). However, with some practical applications, for example one which has a large number of obstacles, this phase is very time-consuming. This paper describes a simple and efficient preprocessing technique to reduce computation time when applying the visibility graph approach for path planning in case of many tiny obstacles. With a large set of randomly generated test problems, the experiment result shows that the computation time gets approximately a reduction factor of one-third.