Vacant House Detection Using Map Data, Images, and Machine Learning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies lack an efficient method to accurately determine the state of vacant houses in a desired district with minimal effort.

Innovation Solution

A vacant house determination device and method utilizing machine learning data and a vacant house probability calculation formula derived from map data, including coordinates, attribute information, and topographic information, to identify vacant houses in captured images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional survey methods are used to determine vacant houses, then measurement precision can be achieved, but productivity is low and requires significant manual effort

Engineering Contradiction:
Improvevacant house determination accuracyVSAvoidsurvey efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual survey methods with an automated system that uses captured images, map data, and machine learning algorithms to determine vacant house status. The determination device automatically processes images and calculates vacant house probabilities without requiring manual inspection of each property.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent creates a virtual representation of the physical environment by capturing images and generating map data that can be processed computationally. This digital copy allows for automated analysis and determination of vacant house status without physical surveying.

Inventive Principle:
Principle #26Copying

2Measurement precision

If comprehensive survey data is collected to improve determination accuracy, then measurement precision improves, but device complexity and operational complexity increase

Engineering Contradiction:
Improvevacant house determination accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The determination device is designed to handle multiple data types (captured images, map data with coordinates, attribute information, topographic information) through a unified processing system. The single device performs functions including image processing, probability calculation, and vacant house determination, reducing the need for multiple specialized systems.

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

3Measurement precision

If manual surveying of each house is performed to ensure accurate identification, then measurement precision is maintained, but loss of time increases significantly

Engineering Contradiction:
Improvevacant house identification accuracyVSAvoidsurvey time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary processing by pre-calculating vacant house probabilities for all houses in the map data before actual determination is needed. This allows the system to quickly reference pre-computed probabilities and make rapid determinations without performing full analysis at the time of querying.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces time-consuming manual house-by-house surveying with automated image processing and probability calculation systems that can analyze multiple properties simultaneously, dramatically reducing the time required for vacant house identification.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3798941B1Vacant house determination device, vacant house determination method, and recording medium
Publication Date: 2025.07.02 NEC CORP
  • EP3798941B1 patent drawingFigure 1
  • EP3798941B1 patent drawingFigure 2
  • EP3798941B1 patent drawingFigure 3

AI summary

A vacant house determination device for grasping the state of vacant houses in a desired district with as little effort as possible is provided. A vacant house probability calculation formula with respect to a predetermined district is derived on the basis of map data that at least includes the coordinates of a vacant house surveyed, the attribute information regarding a structural object, and the topographic information, and the vacant house probability with respect to each of the houses included in the map data is calculated by using the vacant house probability calculation formula. It is determined whether a house having the vacant house probability equal or more than a predetermined value is a vacant house, on the basis of machine learning data for presuming that a house in a captured image is a vacant house and image information regarding a site of a house having a vacant house probability equal to or more than a predetermined value, the machine learning data being generated on the basis of, among pieces of pixel information representing sites of houses in the captured image including a predetermined district, the pixel information representing a site of a house corresponding to the coordinates of a vacant house.