The purpose of this study is to suggest the methodology for finding the optimal positions of surveillance facilities such as closed-circuit televisions (CCTV) or guard booths. Conventionally, the art gallery problem(AGP) to solve the minimum number of guards which together can observe the whole gallery is a well-studied visibility problem in computational geometry. But it is NP-hard problem and on the application of approximate solutions for the problem in real environments, the near solutions have no consideration of visual distance and criminal targets. So, this study suggests the methodology using the visual access and exposure level(VL) which means how far from the limit visibility distance a visual target is. And through the criminal target-based visual exposure($VE_{VL}$) analysis and greedy algorithm, the near optimal surveillance positions guarding the largest targets in short visual range can be obtained. This method was examined and had good results in the real environment such as the inside of exhibition hall and the outside of apartment.