SUMMARYIn this paper, we employ a technique that is used to recognize a material from the reflected ultrasonic waves and propose a category recognition system using two ultrasonic sensors. In this system, the reflected waveforms are handled as two-dimensional images, and which category an object belongs to is instantaneously recognized by pattern matching to reference data by a combinatorial logic circuit. In this system, combinations of information such as the material, angle, and distance of the object are set as the recognition categories. The pattern matching is carried out by a combinatorial logic circuit created directly from the reference data. The recognition experiments with measured waveform data performed according to the proposed method demonstrated the features of the proposed system and its effectiveness when applied to recognition under specific conditions. With five materials set as the categories, the results show that a recognition rate of nearly 100% was obtained, and high recognition results were obtained for almost all of the materials when the categories were extended to combinations of the material, angle, and distance.