Building Plan View Recognition via Room-Based Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing drawing recognition systems struggle to accurately recognize elements in plan views of buildings with different scales, as variations in the number of rooms can lead to decreased recognition accuracy.
Innovation Solution
A drawing recognition system that divides plan views based on the number of rooms, using a recognition model generated through machine learning with sufficient plan views as learning data, and selects a recognition model corresponding to the type of plan view to improve recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a plan view of a building with different scale is recognized directly, then the recognition process is simple, but the recognition accuracy of elements may decrease due to variations in the number of rooms
Solution Approach 1:
The plan view is divided into multiple divided drawings based on the number of rooms included in the plan view. This segmentation allows the recognition system to process each divided drawing with a controlled number of rooms, suppressing variations in room count and improving recognition accuracy without requiring complex preprocessing of the entire plan view at once
2Measurement precision
If the plan view is divided based on the number of rooms, then the recognition accuracy is improved, but the processing complexity increases
Solution Approach 1:
The plan view is divided into multiple divided drawings based on the number of rooms included in the plan view. This segmentation allows the recognition system to process each divided drawing with a controlled number of rooms, suppressing variations in room count and improving recognition accuracy without requiring complex preprocessing of the entire plan view at once
Solution Approach 2:
The recognition system dynamically selects appropriate recognition models based on the type of plan view and processes multiple divided drawings adaptively. This dynamic approach allows the system to handle different building scales and room configurations efficiently, improving accuracy while managing processing complexity through intelligent resource allocation
Data Source
AI summary
To improve recognition accuracy of a plan view of a building. A drawing recognition system 10 includes: an acquisition unit 11 for acquiring drawing data including a plan view of a building; a dividing unit 13 for generating divided drawings by dividing the plan view based on a number of rooms included in the plan view; a recognition unit 14 for generating a recognition result of the plan view by recognizing an element included in the plan view based on the divided drawings; and an output unit 15 for outputting the recognition result of the plan view.