In recent years, AR (Augmented Reality) technology has gradually entered people’s horizons and has become a research hotspot in the field of image processing. At the same time, the mobility and interactivity of AR have also been valued. This paper studies the interaction technology in the AR scene, and proposes a scheme to detect the user’s gaze at the cone perspective. The Leapmotion body sensor was used for gesture extension, and the hardware connection and data transmission scheme of Leapmotion and Holo Lens was designed. Based on the three input modes of sight, gesture and voice, the interaction logic between the user and the system in the AR scene is designed. The network of the Internet of Things is dominated by wireless sensor networks (Wireless Sensor Networks). This paper studies the problem of congestion in wireless sensor network sensing data transmission. WSNs nodes need to transmit the sensed data or preprocessed sensed data to sink nodes. When the data packets (or packets) received by a sensor node exceed its upper limit for forwarding or processing data packets, the excess data packets need to be buffered. When the memory of cached packets is full, excess packets are discarded, causing buffer overflows and causing node-level congestion. This paper proposes an energy balance strategy based on feedback control. The feedback control algorithm is used to adjust the node’s transmit power to realize the redistribution of node load in the network, solve the problem of node energy imbalance in wireless sensor networks, and further extend the sensor network.