Sub-pixel Edge Detection for Construction Object Measurement

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems struggle to accurately measure the shapes of construction materials in complex environments like buildings under construction, due to difficulties in automating the determination of object edges and regions, leading to inefficiencies in time and effort.

Innovation Solution

An object measurement device utilizing a machine learning model to automate edge detection direction and region determination, performing highly accurate edge extraction in units of sub-pixels, using Building Information Modeling (BIM) data for training and virtual observation images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual determination of object regions is performed, then measurement accuracy can be maintained, but time consumption and labor effort increase significantly

Engineering Contradiction:

Improveshape measurement accuracyVSAvoidtime and effort for region specification

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system enables automated self-determination of object regions through machine learning models that automatically identify and segment construction materials in images, eliminating the need for manual region specification while maintaining measurement accuracy. The model processes images autonomously to detect edges and determine object boundaries.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Manual mechanical operations of specifying regions are replaced by an automated image processing system using machine learning. The system substitutes human operators with algorithms that automatically perform edge detection, region segmentation, and object identification through learned patterns from training data.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If automated object recognition systems are used, then time efficiency is improved, but measurement accuracy and recognition precision are insufficient

Engineering Contradiction:

Improveautomation efficiencyVSAvoidconstruction material measurement accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary training of machine learning models using synthetic BIM data and virtual observation images before actual measurement tasks. This pre-training phase prepares the model to accurately recognize construction materials in real-world conditions, ensuring both automation efficiency and measurement precision are achieved simultaneously.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system transitions from two-dimensional image data to three-dimensional shape information by performing edge detection and region determination that extracts spatial dimensions. This dimensional transformation enables accurate measurement of construction materials while maintaining automated processing efficiency.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

3Reliability

If edge detection is performed on entire images, then comprehensive object detection is achieved, but processing time and computational resources increase

Engineering Contradiction:

Improveobject detection completenessVSAvoidedge processing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system segments the image processing task by first identifying candidate regions through initial detection, then performing detailed edge detection only within those specific regions. This segmented approach maintains comprehensive object detection while significantly reducing overall processing time by avoiding unnecessary computation in non-relevant areas.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system applies different processing qualities to different regions: preliminary detection is performed on the entire image to identify potential objects, while high-precision edge detection is applied locally only to identified object regions. This local quality approach ensures detection completeness while optimizing processing efficiency.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20250111518A1Object measurement device and object measurement method

Publication Date: 2025.04.03 TOPCON CORPORATION

AI summary

Provided is an object measurement device for measuring the shape of an object by recognizing the object using a machine learning model, automating determination of an edge detection direction and determination of a region for performing edge processing, and performing highly accurate edge extraction in units of sub-pixels, for example. An object measurement device according to the present invention comprises: an object recognition unit that recognizes an object in an image by inputting an acquired image as input data into a trained model, and outputs a region of the object as a recognition result; an edge detection direction determination unit that determines an edge detection direction for the region of the object recognized by the object recognition unit; an edge processing region determination unit that determines an edge processing region for the region of the object recognized by the object recognition unit; and an edge detection unit that performs edge detection in the edge detection direction determined by the edge detection direction determination unit, for the edge processing region determined by the edge processing region determination unit.