Panoramic Object Detection With Duplicate Filtering for 3D Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D coordinate measurement devices struggle with efficiently detecting and filtering duplicate objects of interest in panoramic images, which is a time-consuming task due to the large amount of data and details in these images.
Innovation Solution
A method and system utilizing a trained machine learning model to detect objects of interest in panoramic images, generating 3D coordinates, and combining them with existing 3D models, while employing frustum filtering and feature matching filtering to identify and remove duplicates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional methods are used to detect objects in panoramic images, then processing completeness is maintained, but processing time and data storage requirements increase significantly
Solution Approach 1:
The system performs preliminary actions by capturing multiple panoramic images at different positions and times before final object detection. These pre-captured images serve as reference data that enables faster and more accurate object detection and filtering in subsequent processing steps, reducing overall processing time while maintaining completeness.
Solution Approach 2:
The patent creates copies of panoramic images and 3D models for comparison and filtering operations. By working with copies rather than the original full-resolution data throughout the entire process, the system reduces processing time and storage requirements while maintaining detection accuracy through comparative analysis of the copied data.
2Reliability
If all detected objects are stored to ensure completeness, then data completeness is improved, but data storage requirements increase
Solution Approach 1:
The system extracts and filters only the essential information needed for accurate object detection and 3D model generation. By extracting key features and filtering out redundant data through comparison of multiple images and frustum analysis, the system reduces storage requirements while maintaining detection accuracy and reliability.
Solution Approach 2:
The patent discards duplicate or redundant object data after verification through frustum filtering and feature matching, then recovers the cleaned, verified object information for final 3D model generation. This process reduces storage requirements by removing duplicates while maintaining completeness through recovery of valid object data.
3Manufacturing precision
If complex filtering methods are applied to reduce duplicates, then data quality is improved, but processing complexity increases
Solution Approach 1:
The filtering process is segmented into distinct stages: frustum filtering based on geometric overlap, feature matching filtering based on visual similarity, and 3D coordinate verification. Each segment handles a specific aspect of duplicate detection, making the overall complex task more manageable and improving 3D model accuracy through systematic processing.
Solution Approach 2:
The system introduces intermediary data structures including frustums (geometric cones representing object fields of view) and feature descriptors as mediators between the raw image data and final 3D models. These intermediaries simplify the filtering process by providing standardized representations that can be efficiently compared and filtered, reducing overall processing complexity while maintaining precision.
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
This approach reduces data storage requirements by effectively detecting and filtering duplicate objects, improving processing efficiency and accuracy in generating revised 3D models.
Implementation Method 1
detecting, using a trained machine learning model, the object of interest in a panoramic image of the environment
Implementation Method 2
determining 3D coordinates for the object of interest
Implementation Method 3
A TOF laser scanner is a scanner in which the distance to a target point is determined based on the speed of light in air between the scanner and a target point
Implementation Method 4
A 3D laser scanner of this type steers a beam of light to a non-cooperative target such as a diffusely scattering surface of an object
Implementation Method 5
combining the 3D coordinates for the object of interest with an existing 3D model of the object of interest to create a revised 3D model
Data Source
AI summary
Examples described herein provide a method for generating a three-dimensional (3D) model of an object of interest using panoramic images of an environment. The method includes detecting, using a trained machine learning model, the object of interest in a panoramic image of the environment. The method further includes determining 3D coordinates for the object of interest. The method further includes combining the 3D coordinates for the object of interest with an existing 3D model of the object of interest to create a revised 3D model of the object of interest.