Multi-Camera 3D Model Generation for Surveillance Depth Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surveillance systems rely on two-dimensional models, leading to inaccuracies in measuring depth and size of targets, and lack the intelligence needed to effectively detect suspicious events, resulting in insufficient security.
Innovation Solution
A method and device that synchronize multiple image capturing devices to generate a 3-D model of an area of interest, allowing for the detection of new elements and improved tracking of objects, using processor-executable instructions to combine images and adjust camera settings for enhanced surveillance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple image capturing devices are synchronized to capture images at intersection points of their field of views, then measurement precision of depth and size is improved, but device complexity increases
Solution Approach 1:
The patent transitions from 2D image analysis to 3D model generation by capturing images from multiple camera viewpoints and synthesizing them into a three-dimensional representation. This dimensional transformation enables accurate depth measurement and size estimation by providing spatial context that single 2D images cannot provide
Solution Approach 2:
The patent combines images from multiple image capturing devices positioned at different locations to create a composite 3D model. By merging data from multiple sources (cameras at different positions), the system achieves comprehensive spatial understanding and improved measurement accuracy that cannot be obtained from a single camera
2Measurement precision
If a 3-D model is generated by combining images from multiple image capturing devices, then detection accuracy of new elements is improved, but loss of time for processing increases
Solution Approach 1:
The system pre-establishes the spatial relationships and calibration parameters between multiple cameras before actual surveillance operations. By performing calibration and coordinate system alignment in advance, the patent reduces the computational burden during real-time operation, allowing faster processing of incoming images for 3D model generation and anomaly detection
3Area of stationary object
If image capturing devices capture images at intersection points of their field of views, then coverage of area of interest is improved, but quantity of images to be processed increases
Solution Approach 1:
The patent focuses computational resources on processing images from specific regions of interest where multiple camera fields of view intersect. Rather than uniformly processing all captured images, the system identifies and prioritizes processing of images from intersection zones where the area of interest is most clearly observed, reducing overall processing load while maintaining comprehensive coverage
Data Source
AI summary
The present disclosure relates to a method for surveillance an area of interest. A surveillance device pre-sets an area of interest for each of plurality of image capturing devices. The surveillance device synchronizes the plurality of image capturing devices to simultaneously observe the area of interest. Upon synchronizing, the surveillance device continuously receives view port details from each of the plurality of image capturing devices. If the received view port details matches with pre-set view port details of the area of interest, then one or more images are captured at the intersection of field of view of plurality of image capturing devices. The captured images are provided to the surveillance device to generate a 3-D model. The 3-D model is observed for detecting presence or absence of a new element in the area of interest.