Measuring equipment with ultrasonic sensor has been developed for the granular surface pattern recognition in a hopper. The estimation algorithm was proposed which utilized the intensity of the received sound and the period in which the wave moved back and forth. The inclination angle and spatial position of the measured point on the surface were determined from the values taking into account the directivities of the sensors, surface condition of the reflection, and attenuation of the sound through the path. The pattern was depicted by microcomputer considering the mutual relations between the locations of the measuring points and the constraint due to the physical properties of the materials treated. The method was applied to the instrument whose resonance frequency and half width were 66kHz and 5 degrees, respectively. The fundamental characteristics of the measuring unit such as frequency dependence, directivities, attenuation, and measurement error have been studied experimentally. Sufficient results were obtained in a model hopper system with sand whose mean diameter was approximately 300 microns for three surface patterns of plane, concave, and convex.