US20250224801
2025-07-10
Physics
G06F3/011
The invention relates to an electronic device designed to support augmented reality (AR) functions. It utilizes a sensor module to gather data about a user's hand, which is then processed by artificial intelligence (AI) models to determine the hand's location and joint shapes. This process helps to identify interactions between the user's hand and objects within an AR image. The device optimizes resource usage by selecting appropriate AI models based on detected interactions, thereby enhancing system efficiency and reducing power consumption.
With the growing demand for AR and mixed reality (MR) services, devices that overlay virtual images onto real-world scenes are becoming increasingly popular. These devices, such as AR glasses, enable users to experience virtual content in their real environments. The advancement of communication technologies like 5G has further propelled the development of such immersive media devices by supporting high-capacity data transmission.
The electronic device integrates a sensor module and a processor to facilitate AR functionalities. It executes applications that provide AR images and utilizes AI models to analyze sensor data about the user's hand. Initially, a first AI model processes this data to detect the hand's position and joint shapes. If an interaction with an AR object is identified, a second AI model is employed for more detailed analysis, thereby adapting resource usage according to the interaction level.
The device is designed to optimize resource consumption by adjusting the computational load based on user interactions. By selecting different AI models depending on interaction detection, the device minimizes unnecessary calculations, thus conserving power and enhancing overall system efficiency. This adaptive approach ensures that resources are utilized effectively only when needed.
The patent also includes various diagrams illustrating the device's components and operational flow. These diagrams demonstrate how image data is processed using different sensors like depth and infrared sensors, as well as camera modules. The illustrations further depict methods for calculating hand positions and joint shapes, as well as scenarios demonstrating user interactions with AR objects. These technical details support a comprehensive understanding of the device's functionality and operational advantages.