Distributed video coding (DVC) is considered as one promising video coding scheme for Wireless Video Sensor Network (WVSN) due to its high compression efficiency and error resilience functionalities, as well as the low encoding complexity. In this paper, an improved DVC scheme based on checkerboard pattern splitting for Wyner-Ziv (WZ) frame is presented. In specific, the modified three-dimensional recursive search (3DRS) motion estimation method is adopted to improve the initial side information (SI) quality at the decoder. Furthermore, the spatio-temporal boundary matching algorithm (STBMA), which utilizes both the spatial and temporal smoothness properties, is used for the SI refinement. The experimental results show that our proposed method can improve the whole performance of DVC codec in transform domain compared to state-of-the-art DVC and be deployed over a real visual sensor platform.