Diesel soot aggregates emitted from a model dynamometer and 11 on-road vehicles were segregated by a micro-orifice uniform deposit impactor (MOUDI). The elemental contents and morphological parameters of the aggregates were then examined by scanning electron microscopy coupled with an energy dispersive spectrometer (SEM-EDS), and combined with a fractional Brownian motion (fBm) processor. Two mode-size distributions of aggregates collected from diesel vehicles were confirmed. Mean mass concentration of 339 mg/m3 (dC/dlogdp) existed in the dominant mode (180-320 nm). A relatively high proportion of these aggregates appeared in PM1, accentuating the relevance regarding adverse health effects. Furthermore, the fBm processor directly parameterized the SEM images of fractal like aggregates and successfully quantified surface texture to extract Hurst coefficients (H) of the aggregates. For aggregates from vehicles equipped with a universal cylinder number, the H value was independent of engine operational conditions. A small H value existed in emitted aggregates from vehicles with a large number of cylinders. This study found that aggregate fractal dimension related to H was in the range of 1.641-1.775, which is in agreement with values reported by previous TEM-based experiments. According to EDS analysis, carbon content ranged in a high level of 30%-50% by weight for diesel soot aggregates. The presence of Na and Mg elements in these sampled aggregates indicated the likelihood that some engine enhancers composed of biofuel or surfactants were commonly used in on-road vehicles in Taiwan. In particular, the morphological H combined with carbon content detection can be useful for characterizing chain-like or cluster diesel soot aggregates in the atmosphere.