Detection of looming obstacles is vitally important to all animals for avoiding predators, conspecifics, and environmental obstacles. The migratory locust, Locusta migratoria, possesses a well-characterized neuron in each optic lobe known as the lobula giant movement detector (LGMD) which integrates visual data into a signal encoding the imminence of collision with an approaching object. While this neuron itself and certain portions of its input network are well-studied from both physiological and modelling perspectives, certain physiological discoveries in recent literature have not yet been reflected in computation models of locust looming detection. Specifically, the posited role of global inhibition in normalizing inputs has not been investigated computationally, and new characterizations of neurons providing feedforward inhibition to the LGMD have not been incorporated into models. Considering this, a model was developed combining features from past literature examples with recent anatomical reconstructions based on neural recordings. This model, consisting of a simulated LGMD neuron and its relevant inputs in the retina, lamina and medulla, will be tested for its ability to replicate features of LGMD responses to more complex looming trajectories which were observed in recent studies – specifically, responses to changes in the velocity of an approaching object. Moreover, the implications of the model regarding the posited physiology of feedforward and global inhibitory elements will be discussed.