3D Building Feature Extraction From DSM Data for Remote Measurement

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems require on-site presence to measure and determine building features, which is inefficient for insurance companies, construction contractors, and real estate sales.

Innovation Solution

A system and method for determining 3D structure from digital surface model (DSM) images using machine learning models, applying object detection and deep learning to outline structures and vegetation, and generating 3D polygons from DSM images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If on-site presence is used to measure and determine building features, then measurement precision is improved, but productivity is worsened due to inefficiency

Engineering Contradiction:

Improvebuilding feature measurement accuracyVSAvoidefficiency of building feature determination

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual on-site measurement systems with an automated computer vision system that processes aerial imagery and DSM data. The system uses machine learning models to automatically detect and measure building features, eliminating the need for physical presence while maintaining measurement accuracy through algorithmic analysis of spatial data.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system creates digital copies of building features by generating 3D models from aerial imagery and digital surface models. These digital representations serve as accurate replicas that can be measured and analyzed remotely, preserving measurement precision while enabling remote operation and significantly improving productivity.

Inventive Principle:

Principle #26Copying

2Productivity

If remote determination methods are used, then productivity is improved, but measurement precision is worsened

Engineering Contradiction:

Improveefficiency of building feature determinationVSAvoidbuilding feature measurement accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent transitions from 2D aerial imagery to 3D digital surface models and three-dimensional building models. By adding the vertical dimension through DSM integration, the system achieves accurate remote measurement of building features including height, volume, and complex geometries that would be difficult to obtain from 2D images alone, thereby maintaining precision while enabling remote operation.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The system employs dynamic machine learning models that adapt to different building types, roof configurations, and vegetation patterns. The models are trained on diverse datasets and can dynamically adjust their detection parameters to maintain high measurement precision across various scenarios, overcoming the limitations of static remote sensing approaches.

Inventive Principle:

Principle #15Dynamics

3Productivity

If machine learning models are applied to determine features from DSM images, then productivity is improved, but device complexity is worsened

Engineering Contradiction:

Improveautomation of building feature determinationVSAvoidcomplexity of machine learning system

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent develops a universal machine learning framework that handles multiple building feature detection tasks (roof facets, walls, vegetation, height measurements) within a single integrated system. This multi-functional approach consolidates what would otherwise require multiple separate systems, managing complexity while providing comprehensive automated analysis of building features from aerial imagery and DSM data.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS12585837B2Determining 3D structure features from DSM data

Publication Date: 2026.03.24 ZESTY AI INC

AI summary

The disclosure includes a system and method for determining a three-dimensional (3D) structure from an image including receiving, using one or more processors, a digital surface model (DSM) image; receiving, using the one or more processors, a non-DSM image; outlining, using the one or more processors, a structure in the non-DSM image; and determining, using the one or more processors, a set of features based on the DSM image.