The warehouse location selection is a processing of selecting allocation center in economic region where there are some supply stations and the certain demand point. Generally, the warehouse location selection model follows some principles: adaptable principle, Coordinative principle, efficient principle and strategic principle [1].
The classical algorithm includes the center-of-gravity approach, the capacitated facility location problem model, the Baumol-Wolfe methods and the P-median selecting location model, in the warehouse location selection. The gravity of the physical system is looked as the best spot in the center-of-gravity approach. Its characteristics is simple computation is the main characteristic, but the accuracy of computation is low. So the result can only be referenced. The capacitated facility location problem model finds the optimal solution by revising the supply scope and moving the demand point for dropping cost in limited network scale situation. The complex computation process is the shortcoming of the model. For example, the optimal solution is showed in first resolving, but it can give the first resolving is the optimal solution, after the model must carry on all circulations. The non-linear problem can be solved in the network layout of Baumol-Wolfe methods and the linear plan is used in every iterating. The optimal plan can been found under some specific constraints. But there are two flaws of the model, the one is that the optimal plan cannot be guaranteed in every conditions, the other is that the constant investment isn’t showed in the model. In order to avoid the above limitation, the method of integrating GIS with remote photo is used in the article.