Greenhouse gas (GHG) emissions from the transportation sector in Korea are currently estimated at an aggregate level, based on vehicle registrations by administrative regions. The sub-classification of detailed vehicle attributes(such as vehicle type, fuel type, and size), which are critical for accurate emission estimation, is also restricted to limited nationwide statistics. To address these limitations, this study proposes a vehicle-type estimation model that integrates optimization techniques with vehicle trajectory data (navigation and DTG), enabling the disaggregation of registration-based statistics into road-level emission inventories. The proposed model estimates traffic volumes by vehicle type using observed counts at monitoring sites and estimated counts at unobserved locations. In addition, digital tachograph (DTG) and navigation trajectory data are utilized to optimize trajectory weights such that vehicle-type-specific target volumes are satisfied while preserving actual driving routes. The resulting traffic volumes are then distributed by vehicle subclass(vehicle type, fuel type, and size) in accordance with vehicle kilometers traveled (VKT) statistics derived from vehicle registration data, allowing the estimation of detailed traffic volumes at the road-link level. Model validation revealed a general tendency to underestimate traffic volumes, with relatively higher errors observed for lower-level road networks and some regions. Nevertheless, overall error levels remained relatively low—ranging from 6–16% by region and 5–21% by road class—demonstrating the robust performance and practical applicability of the proposed methodology across various analytical units. By combining real vehicle trajectory data with optimization techniques, this study establishes disaggregated traffic datasets at the link level by vehicle subclass. These datasets can serve as essential inputs for advanced, high-resolution emission estimation models, thereby enhancing the accuracy of the national GHG inventory system and improving its policy relevance.