In wireless rechargeable sensor networks ( WRSNs ), target coverage is an important issue. However, most studies assumed that the sensors were powered by battery with fixed sensing radiuses. Furthermore, most existing studies applied the Boolean Sensing Model ( BSM ) which cannot reflect the physical features of sensing. As a result, the detection accuracy was decreased in a real application. This paper proposes a target coverage mechanism, called TCSAR , which aims to maximize the surveillance quality of the point of interests ( POIs ) under the constraint of perpetual network lifetime. The proposed TCSAR applies the Probabilistic Sensing Model ( PSM ) and assumes that the sensing radius of each sensor is adjustable. The proposed mechanism consists of two phases. In the scheduling phase, the TCSAR initially evaluates the surveillance contribution of each sensor and then schedules the sensor with the maximal contribution to the bottleneck POIs . In the space time transformation phase, the surveillance quality for the POIs can be further improved by adjusting the sensing radius of some sensors from space dimension to time dimension. The experimental results show that the proposed mechanism outperforms the existing mechanisms in terms of the worst surveillance quality and the quality of monitoring.