Spectral reflectance of objects provides key recognition of objects from remote sensing data. Each surface objects have their own specific spectral reflectance pattern that acts as a spectral fingerprint for object discrimination. This study aims to evaluate the effectiveness of radiometric correction applied to WorldView-2 image by comparing the image correction result with field spectrometer measurement. Some objects were chosen as the basis for observing object spectral reflectance, namely grass, non-mangrove vegetation, mangrove vegetation, soil, and asphalt. The WorldView-2 image was radiometrically and spectrally corrected up to at-surface reflectance level using provided procedures. For reference, the spectral reflectance of selected objects were collected in the field using a JAZ EL-350 field spectrometer (340-1028 nm). In order to perform a direct comparison and evaluation, the results of object spectral reflectance of field spectrometer were resampled based on the center wavelength of WorldView-2 image bands (i.e. from thousand to eight bands). This study found that the spectral reflectance patterns of all targeted objects were similar. However, the most accurate spectral reflectance of WorldView-2 image object was asphalt. Asphalt has high colour homogeneity and relatively stable in a long period of time. This study shows that the standard image radiometric and spectral correction approach is effective to represent the spectral reflectance of objects on Earth surface.