We develop a new method for identifying young stellar objects (YSOs) from star-forming regions using the photometry data from Spitzer's c2d Legacy Project. The aim is to obtain YSO lists as complete as possible for studying statistical properties such as the star formation rate (SFR) and lifetimes of YSOs in different evolutionary stages. The largest obstacle in identifying YSOs comes from background galaxies with similar spectral energy distributions to YSOs. Traditionally, selected color-color and color-magnitude criteria are used to separate YSOs and galaxies. However, since there is no obvious boundary between YSOs and galaxies in color-color diagrams and color-magnitude diagrams (CMDs), those criteria may exclude faint YSOs near the boundary. In this paper, we separate the YSOs and galaxies in a multi-dimensional (multi-D) magnitude space, which is equivalent to using all variations of CMDs simultaneously. Comparing sources from molecular clouds to Spitzer's SWIRE data, which have a negligible amount of YSOs, we can naturally identify YSO candidates (YSOc) located outside of the galaxy-populated regions in the multi-D space. In the five c2d surveyed clouds, we select 322 new YSOc and miss/exclude 33 YSOc compared to Evans et al. (2009, Cat. J/ApJS/181/321), and this results in 1313 YSOc in total. As a result, SFR increases 28% correspondingly, but the lifetimes of YSOs in different evolutionary stages remain unchanged. Compared to theories by Krumholz & McKee (2005ApJ...630..250K), our derived SFR suggests that star formation at a large scale is dominated by supersonic turbulence rather than magnetic fields. Furthermore, we identify seven new very low luminosity objects.