Live View Ghost Object Removal Using Motion-Based Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Smartphone cameras often suffer from lens flare phenomena like camera ghosting, which lowers image quality and requires manual post-processing to remove ghost objects, and existing methods struggle to automatically detect and remove ghost objects in real-time, especially when multiple light sources are present.
Innovation Solution
An electronic apparatus equipped with a camera, display, sensor, and processor that identifies areas of predetermined brightness, tracks their motion, and compares it with the apparatus' motion to automatically detect and remove ghost objects in live view images using image processing techniques, including neural networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual post-processing is used to remove ghost objects, then image quality can be improved, but time consumption and operational complexity increase
Solution Approach 1:
The system performs preliminary detection of ghost objects in the live view image before final image capture. By identifying ghost objects in advance using brightness threshold detection and motion analysis, the system prepares removal parameters ahead of time, enabling automatic removal during post-processing without requiring manual intervention during the actual editing phase.
Solution Approach 2:
The system implements self-service by automatically detecting, tracking, and removing ghost objects without user intervention. The processor autonomously analyzes live view images, identifies ghost objects through motion comparison, and applies removal algorithms to both live view and captured images, eliminating the need for manual post-processing operations.
2Productivity
If automatic detection methods are used to identify ghost objects, then productivity is improved, but detection precision may deteriorate when multiple light sources are present
Solution Approach 1:
The system segments the detection process into distinct stages: first identifying multiple high-brightness areas in the live view image, then individually tracking the motion of each area separately. By comparing the motion of each brightness area independently against the device's actual motion, the system can accurately distinguish ghost objects from real light sources even when multiple light sources are present in the scene.
Solution Approach 2:
The system employs feedback mechanisms by continuously monitoring the motion of detected brightness areas and comparing it with the device's motion data from the sensor. This feedback loop allows the system to dynamically adjust its detection criteria and accurately identify ghost objects based on motion inconsistency, maintaining high detection precision in complex scenes with multiple light sources.
3Loss of time
If real-time processing is implemented to remove ghost objects, then time loss is reduced, but device complexity increases
Solution Approach 1:
The system performs preliminary detection and tracking of ghost objects during the live view phase before the actual photo capture. By completing the identification and parameter preparation in advance, the actual removal process during post-processing becomes a simple parameter application task, reducing real-time processing demands while maintaining real-time visual feedback for the user.
4Measurement precision
If motion comparison with sensor data is used to identify ghost objects, then detection precision is improved, but device complexity increases
Solution Approach 1:
The system uses the device's motion sensor as an intermediary to obtain accurate motion reference data. By comparing the motion of detected brightness areas against the sensor's measurement of actual device motion, the system achieves high-precision ghost object identification. The sensor acts as a mediator that provides ground-truth motion data, simplifying the comparison logic while maintaining high detection accuracy.
Data Source
AI summary
An electronic apparatus includes a camera, a display, a sensor, and at least one processor configured to: control the display to display a plurality of first live view images obtained through the camera; identify a first area and a second area, each having a predetermined brightness or greater, in the plurality of first live view images; identify one area from among the first area and the second area as an area subject to removal by comparing a first motion of the first area and a second motion of the second area with a third motion of the electronic apparatus obtained through the sensor; perform an image processing on an area corresponding to the area subject to removal in a plurality of second live view images, obtained after the plurality of first live view images; and control the display to display the image processed plurality of second live view images.