Indoor 3D Mapping Using Depth Priors Instead of LiDAR

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing indoor mapping systems face challenges such as high costs, power demands, bulkiness, and complexity due to reliance on LiDAR sensors, and struggle with occlusions and incomplete data capture using depth cameras, while integration of architectural blueprints requires sophisticated algorithms and computational resources beyond real-time capabilities.

Innovation Solution

A system utilizing RGB cameras and self-supervised learning, combined with geometric and semantic cues, generates a depth prior to refine a depth prediction model, enabling real-time construction of a continuous three-dimensional map by leveraging lightweight sensors and pseudo supervision for adaptability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If LiDAR sensors are used for depth measurement, then measurement precision is improved, but device complexity and cost increase

Engineering Contradiction:
Improvedepth measurement accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent uses depth camera data to create a simplified copy of the depth information that LiDAR would provide, processing this copied data through the monocular depth estimation model to achieve accurate depth measurements without requiring actual LiDAR hardware

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces expensive, complex LiDAR sensors with cheaper depth cameras and computational models, using software-based depth estimation as a disposable alternative to expensive hardware solutions

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

2Measurement precision

If multiple sensors are combined to improve coverage and accuracy, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improvedepth measurement accuracyVSAvoidsensor system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges depth camera data with monocular depth estimation model predictions into a unified depth map, combining multiple information sources computationally rather than physically combining multiple sensors

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The system uses a single depth camera that serves multiple functions: capturing visual imagery, providing depth data, and serving as input for the monocular depth estimation model, eliminating the need for separate specialized sensors

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Manufacturing precision

If architectural blueprints are integrated into the mapping system, then manufacturing precision is improved, but device complexity increases

Engineering Contradiction:
Improvelayout reconstruction accuracyVSAvoidalgorithm complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary alignment between the captured environment and architectural blueprints before final mapping, pre-processing the blueprint data to match the captured spatial coordinates and scale, thereby simplifying subsequent integration steps

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS20260073633A1Optimizing environment mapping with depth prediction
Publication Date: 2026.03.12 FIELD AI INC
  • US20260073633A1 patent drawing
  • US20260073633A1 patent drawing
  • US20260073633A1 patent drawing

AI summary

A method of and system for generating a three-dimensional map of an environment can include obtaining a first visual data set, generating a depth prior based on the first visual data set, refining a depth prediction model based on the depth prior, generating a layout based on a refined depth prediction model, and constructing a continuous three-dimensional map of the environment based on the layout and aggregated depth measurements. The visual data set can include visual imagery data and depth data. The depth prior can include geometric cues and semantic cues