Estimating the mass of a meteoroid passing through the Earth's atmosphere is essential to determining potential meteorite fall positions. High-resolution fireball images from dedicated camera networks provide the position and timing for fireball bright flight trajectories. There are two established mass determination methods: the photometric and the dynamic. A new approach is proposed, based on the dynamic method. A dynamic optimization initially constrains unknown meteoroid characteristics which are then used in a parametric model for an extended Kalman filter. The extended Kalman filter estimates the position, velocity, and mass of the meteoroid body throughout its flight, and quantitatively models uncertainties. Uncertainties have not previously been modeled so explicitly and are essential for determining fall distributions for potential meteorites. This two-step method aims to automate the process of mass determination for application to any trajectory data set and has been applied to observations of the Bunburra Rockhole fireball. The new method naturally handles noisy raw data. Initial and terminal bright flight mass results are consistent with other works based on the established photometric method and cosmic ray analysis. A full analysis of fragmentation and the variability in the heat-transfer coefficient will be explored in future versions of the model.