Multi-view Imaging System Using Mirroring for Single Camera Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-view imaging systems require multiple cameras or sequential captures, leading to time-consuming and resource-intensive processes, with issues in capturing coherent three-dimensional information due to changing lighting conditions and inability to detect mutual occlusions between mirrored images.
Innovation Solution
A method and device using a single camera with a mirroring system to generate multiple virtual views by computing mirroring sequences for each image pixel, allowing mapping to virtual views and projecting light beams effectively, while accounting for the geometry of mirrors and object characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple cameras are used to capture multi-view images simultaneously, then the depth information and three-dimensional properties can be captured, but the device complexity and resource consumption increase
Solution Approach 1:
The patent uses mirrors to create virtual copies of the camera's viewpoint. By placing mirrors around the object, the single camera captures multiple virtual views that appear as if taken from different positions. This copying approach through optical reflection achieves multi-view capture without requiring multiple physical cameras, thus reducing device complexity while maintaining the ability to extract three-dimensional information
2Measurement precision
If sequential image captures are used to generate multiple views, then the three-dimensional information can be extracted, but the time consumption increases and lighting conditions may change
Solution Approach 1:
The patent pre-arranges mirrors in specific positions around the object before capture. This preliminary setup creates a stable optical path that reflects light from the object to the camera simultaneously for all virtual views. By preparing the mirroring configuration in advance, the system captures all multi-view information in a single exposure, eliminating the time delays and lighting condition changes associated with sequential captures
3Productivity
If a simple mirroring setup is used to generate virtual views, then the capture time is reduced, but the ability to detect mutual occlusions and exploit multi-view information is limited
Solution Approach 1:
The patent segments the captured image into multiple distinct regions, each corresponding to a different virtual view. By analyzing the geometric relationships and occlusion patterns within each segmented region, the system can detect mutual occlusions between different virtual viewpoints. This segmentation approach transforms a complex single-image analysis problem into manageable regional analyses, enabling accurate occlusion detection while maintaining fast single-shot capture
4Reliability
If multiple cameras are deployed, then simultaneous multi-view capture is achieved, but the resource consumption and system complexity increase
Solution Approach 1:
The patent makes a single camera multi-functional by using mirrors to create multiple virtual viewpoints from one physical device. The single camera performs the work of multiple cameras by capturing reflected images that represent different spatial positions. This universal approach maintains lighting condition consistency across all views since they are captured simultaneously by the same sensor, while avoiding the resource consumption of deploying multiple cameras
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables the capture of dense, calibrated multi-view data from a single image, achieving perfect synchronization and uniform radiometric and colorimetric properties across views, overcoming limitations of prior art by allowing dynamic scene recording and efficient use of sensor area.
Implementation Method 1
a mirroring system (040) having one or more mirrors, each mirror providing a reflection of the object, each reflection corresponding to a virtual view of the object
Data Source
Figure 1(a)~2(b)
Figure 3(a)~4
Figure 5~6
AI summary
According to the present invention, there is provided a method and device for identifying multiple views of an object in a single image that is taken by a camera using a generic mirroring system. This allows for example the use of the multi-view information, which is included in the image, in multi-view image reconstruction algorithms. The method straightforwardly extends to a projection application wherein a projector is able to selectively project onto different locations of an object, from a single point of projection.