AR Image Processing Dynamic Focus Selection

The patent implements dynamic scene composition adjustment by detecting the addition of virtual objects and automatically reconfiguring the focusing targets. The system transitions from a static composition to a dynamic one where the focusing object is selected based on the current scene contents, including both virtual and real objects. This allows the composition to adapt automatically to maintain optimal focusing effects despite the presence of multiple objects.

The patent changes the parameter of focusing target selection by introducing a selection mechanism that considers both virtual objects and real objects in the scene. The system adjusts the focusing parameters dynamically by determining which object should be the focusing target based on pre-set conditions, such as object importance, position, or user preferences, thereby optimizing the scene composition.

The system dynamically adjusts the focusing effect by selecting appropriate focusing targets from the available objects in the scene. When virtual objects are added, the system re-evaluates the scene composition and automatically selects the most suitable object for focusing, ensuring that the focusing effect remains optimal despite the added complexity of AR elements.

The patent implements a feedback mechanism where the system continuously monitors the shooting scene and detects when virtual objects are added. Based on this detection, the system automatically adjusts the focusing parameters and selects new focusing targets, creating a closed-loop control system that maintains optimal focusing effects in response to changing scene conditions.

The patent applies preliminary action by performing focusing and blurring operations on the captured image before the final synthesis is completed. The system pre-processes the image by identifying and processing the focusing target and blurring other regions, then integrates this processed image with the virtual object. This preliminary processing ensures high image quality while maintaining efficient processing speeds.

The patent segments the image processing into distinct operations: detecting virtual objects, selecting focusing targets, processing the captured image with focusing and blurring, and finally synthesizing the results. This segmentation allows each operation to be optimized independently, maintaining overall processing speed while improving image quality through dedicated focus on the focusing target.