| CPC G06V 40/10 (2022.01) [G06T 2207/10 (2013.01); G06V 40/171 (2022.01)] | 18 Claims |

|
1. A method implemented by an electronic device wherein the method comprises:
obtaining video frames comprising a face of a user and a hand of the user;
recognizing a hand action of the user relative to the face in the video frames;
determining, in response to recognizing the hand action, a target hand action;
determining, from the at least one of the video frames, a first intersection region between a first region to which a finger of the user points in the target hand action and a second region in which the face is located;
determining a detection target in the first intersection region;
determining a geometric shape of the first region using a fingertip of the finger as a reference point and a direction to which the finger points as a reference direction, wherein the geometric shape has a size and a contour that are preset by the user;
determining, from an extended content library and based on the detection target and a shape of the detection target, extended content associated with the detection target and the shape;
outputting the extended content;
obtaining a real-time image of the user using a camera of the electronic device;
displaying the real-time image in a first interface of the electronic device; and
obtaining, from the real-time image, a video frame that comprises both the face and the hand.
|