Space-based direct imaging missions (HabEx, LUVOIR) would observe reflected light from exoplanets in the habitable zones of Sun-like stars. The ultimate---but not sole---goal of these concept missions is to characterize such planets. Knowing an exoplanet's orbit would help two-fold: (i) its semi-major axis informs whether the planet might harbour surface liquid water, making it a priority target; and (ii) predicting the planet's future location would tell us where and when to look. The science yields of HabEx and LUVOIR depend on the number, cadence, and precision of observations required to establish a planet's orbit. We produce mock observations using realistic distributions for the six Keplerian orbital parameters, experimenting with both Beta and uniform eccentricity distributions, and accounting for imperfect astrometry ($\sigma$ = 3.5 mas) and obscuration due to the inner working angle of a high-contrast imaging system (IWA = 31 mas). Using Markov chain Monte Carlo methods, we fit the orbital parameters, and retrieve their average precisions and accuracies as functions of cadence, number of epochs, and distance to target. Given the time at which it was acquired, each image provides two data: the $x$ and $y$ position of the planet with respect to its star. Parameter retrieval based on one or two images is formally under-constrained; yet the semi-major axis posterior can be obtained semi-analytically. For a planet at 1 AU around a star at a distance of 10 pc, three epochs constrain the semi-major axis to within $\lesssim$5\%, if each image is taken at least 90 days apart.

Direct imaging is likely the best way to characterize the atmospheres of Earth-sized exoplanets in the habitable zone of Sun-like stars. Previously, Stark et al. (2014, 2015, 2016) estimated the Earth twin yield of future direct imaging missions, such as LUVOIR and HabEx. We extend this analysis to other types of planets, which will act as false positives for Earth twins. We define an Earth twin as any exoplanet within half an e-folding of 1 AU in semi-major axis and 1 $R_{\oplus}$ in planetary radius, orbiting a G-dwarf. Using Monte Carlo analyses, we quantify the biases and planetary false positive rates of Earth searches. That is, given a pale dot at the correct projected separation and brightness to be a candidate Earth, what are the odds that it is, in fact, an Earth twin? Our notional telescope has a diameter of 10 m, an inner working angle of 3{\lambda}/D, and an outer working angle of 10{\lambda}/D (62 mas and 206 mas at 1.0 {\mu}m). With no precursor knowledge and one visit per star, 77% of detected candidate Earths are actually un-Earths; their mean radius is 2.3 $R_{\oplus}$, a sub-Neptune. The odds improve if we image every planet at its optimal orbital phase, either by relying on precursor knowledge, or by performing multi-epoch direct imaging. In such a targeted search, 47% of detected Earth twin candidates are false positives, and they have a mean radius of 1.7 $R_{\oplus}$. The false positive rate is insensitive to stellar spectral type and the assumption of circular orbits.