It is of paramount importance to obtain accurate node positions in ad hoc networks for various location-based services. Hop-count based localization provides a low-cost and efficient approach to yield the node locations with moderate accuracy. In this paper, an enhanced hop-count based localization algorithm, Segmentation-aided and Density-aware Hop-count (SDH), is proposed to improve the estimation accuracy. In the proposed scheme, the path from one node to the reference nodes (RNs) is first divided into several three-hop segments; each segment distance can be estimated by density-aware hop- count localization. Based on the proposed SDH localization algorithm, we further propose a relative location estimation method for one mobile node to reach the interested node without knowing the location map of whole network. The advantages of the proposed scheme over other conventional schemes are verified by numerical analysis and simulation.