Multi-Camera Object Tracking with Seamless Auto Framing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
When using multiple cameras, it is challenging to naturally track an object and maintain seamless auto framing, as screen display can become discontinuous due to differences in field of view and zoom operations, leading to unnatural screen switching and degraded image quality.
Innovation Solution
An electronic device with multiple cameras and a processor that switches between cameras based on object movement, adjusting magnification and field of view to ensure continuous tracking and auto framing, using a method that includes obtaining images from multiple cameras, determining the appropriate magnification based on object position, and displaying the images in a seamless preview.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple cameras are used to capture images, then the field of view and zoom capability are improved, but the screen display becomes discontinuous and tracking becomes unnatural
Solution Approach 1:
The system continuously monitors object position and movement across camera feeds, using feedback loops to detect when an object moves out of the current camera's optimal viewing area. This triggers automatic camera switching and reframing actions to maintain continuous tracking of the selected object across multiple cameras without disrupting the user experience.
Solution Approach 2:
The patent implements dynamic camera switching and adaptive reframing where the system automatically adjusts which camera is active and modifies the field of view based on real-time object position and movement. This dynamic adaptation ensures seamless transitions between cameras while maintaining natural-looking continuous tracking of the selected object.
2Measurement precision
If zoom operation is applied to display object in designated size, then the object tracking precision is improved, but the background framing becomes limited and image quality degrades
Solution Approach 1:
The system dynamically adjusts zoom levels and field of view based on the selected object's position and the current camera's capabilities. When an object is detected, the system automatically determines the appropriate zoom level to display the object in a designated size while maintaining adequate background context, preventing excessive zoom that would degrade image quality.
Solution Approach 2:
The patent changes imaging parameters such as zoom level, field of view, and magnification based on the detected object's characteristics and position. The system selects optimal parameter combinations that balance object detection precision with background framing requirements, avoiding excessive zoom while ensuring the object is displayed at the designated size.
Data Source
AI summary
An electronic device, and method performed by the electronic device, is provided. The electronic device includes multiple cameras, a display, and at least one processor configured to execute an application supporting image capturing using the multiple cameras, obtain a first image having a first field of view via a first camera among the multiple cameras, display a preview using the first image on the display in a state where a magnification for the image capturing is configured to be a first magnification, change the preview using the first image, displayed on the display, to include at least one object identified in the first image according to a movement of the at least one object, change the magnification for the image capturing to a second magnification, based on the at least one object being positioned in a designated region of the first field of view, obtain a second image having a second field of view via a second camera among the multiple cameras, the second magnification being determined based on the first field of view, the second field of view, and a position of the at least one object in the first field of view, and display at least a part of the second image on the display as the preview, in a state in which the magnification for the image capturing is configured to be the second magnification.