In this paper, we propose a robust and real-time localization method for dynamic environments based on a sensor network; the method combines landmark image information obtained from an ordinary camera and distance information obtained from sensor nodes in an indoor environment. The sensor network provides an effective method for a mobile robot to adapt to changes and guides it across a geographical network area. To enhance the performance, we used a charge-coupled device (CCD) camera and artificial landmarks for self-localization. Experimental results showed that global localization can be achieved with high robustness and accuracy using the proposed localization method.