Indoor Dimensional Measurement Using Single Depth Image Plane Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining dimensions in indoor scenes, such as those used in building construction and robotics, face challenges with high costs and low accuracy due to the use of expensive 3D sensors with long scanning times or short-range sensors that accumulate drift errors in multi-frame registration.
Innovation Solution
A method utilizing a single depth image from a depth sensor, combined with user-guided repositioning for high-quality data collection, enables efficient and accurate dimensional analysis by extracting planes and performing geometric analysis, thereby avoiding error accumulation and providing real-time estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 3D sensors with rotating laser are used to generate 3D models, then measurement precision and coverage are improved, but device cost and scanning time increase
Solution Approach 1:
The patent uses inexpensive depth sensors (similar to consumer 3D cameras) instead of expensive rotating laser 3D sensors. These cheaper sensors have shorter range and smaller field of view but sufficient for indoor dimensional measurements when used strategically from multiple positions
Solution Approach 2:
The patent divides the indoor scene into multiple measurement zones and uses multiple depth sensor positions to cover the entire space. Instead of using one expensive sensor to capture everything at once, the system uses multiple inexpensive sensors positioned at different locations to segment the measurement task
2Area of stationary object
If multiple frames are registered using SLAM techniques to generate large-scale 3D models, then coverage area is improved, but measurement precision deteriorates due to drift errors
Solution Approach 1:
The patent extracts and uses only the essential depth information from each individual frame rather than relying on complex multi-frame registration. By taking out the critical depth data from single frames and using geometric analysis directly on these extracted data, the system avoids the drift errors that accumulate during SLAM registration while still achieving comprehensive coverage through multiple positioned sensors
3Productivity
If a single depth image is used for dimensional analysis, then computational efficiency is improved, but measurement precision may be insufficient
Solution Approach 1:
The patent performs preliminary geometric analysis and plane extraction on the single depth image before any dimensional measurement is attempted. By preparing the data structure and identifying relevant geometric elements in advance, the system maximizes the information extracted from the single frame, ensuring sufficient measurement precision without requiring multiple frames or complex registration
Data Source
AI summary
A method determines dimensions in a scene by first acquiring a depth image of the scene acquired by a sensor, and extracting planes from the depth image. Topological relationships of the planes are determined. The dimensions are determined based on the planes and the topological relationships. A quality of the dimensions is evaluated using a scene type, and if the quality is sufficient outputting the dimensions, and otherwise outputting a guidance to reposition the sensor.