Recent advances in the live imaging of embryonic development promise to revolutionise our understanding of morphogenetic processes. Tissue mechanics are a vital link between cellular protein expression and the changing shapes of embryos, but remain one of the least well-understood aspects of development. As a computational biologist and image analyst, I develop methods to analyse in vivo morphogenetic movements and to infer biomechanical parameters. This approach can be broken down into five steps. First, 4D imaging datasets are translated into cell trajectories and the evolution of cell shapes over time (Blanchard et al., 2009, Nat. Methods). Second, local tissue deformation rates are quantified and broken down into the additive contributions of different cell behaviours (Blanchard, 2017, Phil. Trans. Roy. Soc. B). Third, the fluorescence intensity of tagged Myosin motors are quantified as a proxy for cell contractility. Fourth, the above information is combined to estimate mechanical parameters in vivo (Machado et al., 2015, BMC Biol.). Finally, we test our understanding with computational models (Dicko et al., 2017, PLoS Comp. Biol.). I will summarise recent progress in these five areas, applied to epithelia in the Drosophila embryo. We find that trans-tissue cables at specific cell-cell junctions drive convergence of the germ-band (Tetley*, Blanchard* et al., 2016, eLife). In recent work we show that trans-tissue stress chains across the apico-medial cortices of cells can also drive tissue contraction.