A class of Wide-Field Neurons (WFNs) is prevalent in the brains of animals such as birds, mammals, and insects. These WFNs prefer small moving objects, and their single dendrite is sufficient to drive the somatic response under certain conditions. Compared to the classical weighted sum neuron models, WFNs are more suitable for the detection of local small stimuli. Furthermore, a preference for long-lasting continuous movement targets, i.e., the facilitation mechanism, is a common attribute of WFNs. Based on the above phenomena and the demand for animals to realize a highly successful capture rate in cluttered natural environments, this paper proposes the hypothesis that the dendrite field of WFNs performs dynamic programming (DP), the optimal solution from a Bayesian perspective to a tracking problem if the state space is discrete, and two supporting modeling criteria are proposed. Based on the hypothesis that WFNs’ dendrite field performs DP, this paper designs a coding model of WFNs neuron population for motional target detection in noise background, and verifies the validity of the model.