Automated Image Location Determination Using Room Shape Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods face challenges in effectively capturing and representing building interior information, particularly in determining the acquisition locations of images within buildings without physical presence, and in accurately scaling and populating floor plans with visual data for remote users.
Innovation Solution
The use of computing devices to automatically determine image acquisition locations based on room shapes and features, employing techniques like neural networks for 3D room shape estimation and matching, without requiring depth sensors or distance-measuring devices, to generate and present image acquisition location information for improved navigation and visualization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If automated image acquisition location determination is implemented without depth sensors, then device complexity is reduced, but measurement precision of acquisition locations may be compromised
Solution Approach 1:
The patent replaces depth sensors and distance-measuring devices with a computational approach using 2D images and neural networks. The system substitutes mechanical/optical measurement hardware with software-based 3D room shape estimation and image matching algorithms, eliminating the need for specialized sensing hardware while maintaining acquisition location determination capability.
Solution Approach 2:
The patent introduces determined room shapes as an intermediary representation between 2D images and acquisition locations. By creating a 3D geometric model of the room from 2D images and using it to match and locate images, the system bridges the gap between simple image capture and precise location determination without requiring direct depth measurement hardware.
2Measurement precision
If floor plans are manually constructed and maintained, then accuracy of building information can be ensured, but productivity and time consumption increase significantly
Solution Approach 1:
The patent enables automated generation and updating of floor plans using images captured during normal navigation. The system automatically processes images, determines acquisition locations, and updates building information without requiring manual intervention for each update, allowing the system to self-maintain accurate building representations.
Solution Approach 2:
The patent performs preliminary automated processing of images to extract building information before it is needed for navigation or display. By pre-processing images to determine room shapes, acquisition locations, and floor plan updates, the system prepares building data in advance, reducing the need for manual construction and maintenance efforts when the information is actually needed.
3Loss of information
If detailed building interior information is captured for remote viewing, then information completeness improves, but difficulty of representing and displaying the information increases
Solution Approach 1:
The patent transforms 2D images into 3D room shape representations and maps them to 2D floor plans with associated metadata. By adding the dimension of spatial location information (acquisition location on floor plan, orientation angles) to the visual data, the system enables comprehensive building information representation while maintaining manageable data structures that can be efficiently stored and displayed.
Data Source
AI summary
Techniques are described for computing devices to perform automated operations to determine the acquisition locations of images, such as within a building interior based on automatically determined shapes of rooms of the building, and for using the determined image acquisition location information in further automated manners. The image may be a panorama image or of another type (e.g., a rectilinear perspective image) and acquired at an acquisition location in a multi-room building's interior, and the determined acquisition location for such an image may be at least a location on the building's floor plan and optionally an orientation/direction for at least a part of the image—in addition, the automated image acquisition location determination may be further performed without having or using information from any depth sensors or other distance-measuring devices about distances from an image's acquisition location to walls or other objects in the surrounding building.