Conventional single pixel evaluation (SPE) significantly improves the spatial resolution of PIV measurements to the physical limit of a CCD camera based on the forward difference interrogation (FDI). This paper further enhances the computational algorithm to second-order accuracy by simply modifying the numerical scheme with the central difference interrogation (CDI). The proposed central difference scheme basically superposes the forward-time and the backward-time correlation domains, thus resulting in reduced bias error as well as rapid background noise elimination. An assessment of the CDI SPE algorithm regarding the measurement errors was achieved via numerous synthetic images subject to a four-roll mill flow. In addition, preliminary wall shear stress (WSS) measurements regarding different algorithms are also evaluated with an analytical turbulent boundary flow. CDI scheme showed a 0.32% error deviated from the analytical solution and improved the same error in FFT-based correlation correlation (FFT-CC) by 32.35%. To demonstrate the performance in practice, in-vitro measurements were implemented in a serpentine microchannel made of polydimethyl siloxane (PDMS) for both CDI SPE and spatial cross-correlation. A series of steady-state flow images at five specified regions of interest were acquired using micro-PIV system. Final comparisons of the WSS regarding the Pearson correlation coefficient, R2, between the numerical schemes and the simulations showed that an overall result was improved by CDI SPE due to the fine resolution and the enhanced accuracy.