A study on the application of artificial intelligence in the design of intelligent medical robots

Abstract

Abstract With the development of artificial intelligence and robotics technology, the combination of artificial intelligence and medical device research and development has been promoted, which is an important product of the development of artificial intelligence. In this paper, the general structure of the intelligent medical robot is designed by combining artificial intelligence technology and robotics-related technology. Then, the binocular vision function of the robot was realized by visually acquiring the image of the target object, 3D reconstruction of the target object, and combining the SIFT image recognition algorithm and target tracking algorithm. Then, a new speech recognition algorithm was constructed to realize the human-robot interaction function with the medical robot based on the deep learning Transforme network after the construction of the human acoustic model. Finally, the designed intelligent medical robot was tested, and its overall performance was evaluated. The results show that the recognition errors of the intelligent medical robot on the features of the items are all within 0.05, the recognition errors on the features of the human body are within 0.2, and the speed of the target tracking is between 6km/h and 16km/h. The average recognition accuracy of the medical robot for voice commands is about 0.9, the recognition time is about 0.7s, the normal working rate of each function is more than 0.99, and the test speed is within 2s.