Detection of ice on the skin surface of aircraft in the stationary state is a challenging task. In this study, an ultrasonic guide detection technique is proposed to distinguish among three ice types (glaze ice, mixed ice, and rime ice on plate-like structure surfaces) and determine the ice thickness by extracting the characteristic parameters of the time and frequency domains. First, the S0 mode signal of the ultrasonic Lamb wave is acquired using ABAQUS simulation software. The degree of S0 mode amplitude attenuation influenced by the ice thickness and type is observed in the time domain. In the frequency domain, the ice coefficient is introduced to realize nonlinear mapping on the ice thickness and type. Subsequently, experimental verification is performed on the three ice thickness types with Lamb waves. A comparison with the simulation results indicates that the method is feasible. Finally, a method of analyzing the type and thickness of ice is described.