This research was conducted to verify whether the semi-automatic method of air-void system characterization (using a flatbed scanner) could be used in lieu of the manual (ASTM C457) method. This research objective was accompanied by evaluating the critical air-void system parameters required to ensure freeze-thaw resistance of ternary concrete containing ordinary portland cement (OPC), 20 % of fly ash (FA), and 5 % of silica fume (SF). It was observed that, due to adverse effect of high superplasticizer dosage on air-void system quality (particularly on specific surface), the majority of high slump ternary mixtures exhibited poor freeze-thaw resistance, even if spacing factor was below the 0.20 mm recommended value. However, when the ternary mixtures were prepared with slump below the limit commonly specified for bridge deck applications (190 mm), the critical air-void system parameters of the OPC/FA/SF concrete were comparable to those typically associated with adequate quality air-void system in conventional concrete. Furthermore, since the critical value of specific surface was independent of slump (i.e., superplasticizer dosage), specific surface appeared to be more objective and reliable predictor of freeze-thaw resistance than spacing factor. When comparing the semi-automatic and manual methods of air-void system characterization, a reasonably strong agreement was obtained for such parameters as air content, void frequency, and spacing factor. Although the results of these methods did not match perfectly, the data suggested that the flatbed scanner technique could be efficiently used for discerning between freeze-thaw durable and non-durable concretes.