Recently, unmanned aerial vehicles (UAVs) configured to provide real-time data for emergency rescue have been integrated into vehicular ad hoc networks (VANETs) to establish UAV-assisted VANETs. Because the communication channels between vehicles and UAVs are open and insecure, an authentication and key agreement (AKA) scheme is needed to secure such channels. Furthermore, a two-factor authentication mechanism is essential. However, most existing two-factor AKA schemes cannot achieve truly two-factor security while resisting password guessing attacks effectively. To address this problem, we propose a lightweight, provably secure, and practical two-factor AKA scheme for UAV-assisted VANETs based on chaotic maps by adopting techniques including fuzzy verifiers and honeywords. In this study, considering the limited computational storage of vehicles and UAVs, a hybrid generation algorithm is proposed to optimize the storage of honeywords. Through rigorous security proofs and detailed security analyses, we demonstrate that the scheme meets the security requirements of VANETs and can resist more security attacks. By utilizing the JPBC library to performance evaluation, a comprehensive comparison with related schemes reveals that the proposed scheme achieves a better balance between communication and computational overhead.