Being deployed in narrow but long area, strip wireless sensor networks (SWSNs) have drawn much attention in applications such as coal mines, pipeline and structure monitoring. One of typical characteristics of SWSNs is the large hop counts, which leads to long end-to-end delivery delay in low-duty-cycle SWSNs. To reduce the delay, pipeline scheduling is a promising technique, which assigns sensor nodes sequential active time slots along the data forwarding path. However, pipeline scheduling is prone to failure when communication links are unreliable. In this paper, we propose an opportunistic pipeline scheduling algorithm (OPS) for SWSNs, based on the observation that sensor nodes in SWSNs can overhear data transmissions passing by them. OPS exploits nodes outside the data forwarding path to opportunistically provide links when transmission failure happens, and hence maintains the pipeline forwarding instead of retransmission in the next duty cycle. Theoretical calculation shows that the expectation delay of OPS is always smaller than that of existing methods when the link quality is <100 %. Both extensive simulations and experiments are conducted. The results verify that the average end-to-end delivery delay of OPS is usually <60 % of that of existing methods, while the energy cost is almost the same.