Exercise Route Recommendation Using Satellite Image Clustering

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing exercise route recommendation applications do not accurately reflect the actual characteristics and conditions of exercise routes, as they rely solely on constructed map data and fail to consider real-time environmental and shape features.

Innovation Solution

A method that extracts candidate exercise routes from satellite image data, analyzes shape and environmental characteristics, clusters these routes using a convolutional neural network, and provides personalized recommendations based on user preferences and exercise mode, incorporating features like architectural structures, natural environments, and route conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If exercise routes are recommended based on constructed map data, then the recommendation process is simple and fast, but the accuracy of reflecting actual route characteristics and conditions is poor

Engineering Contradiction:

Improveaccuracy of route characteristicsVSAvoidcomplexity of recommendation system

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary extraction of route characteristics and environmental features from satellite images before recommendation. Candidate areas are identified and pre-processed offline, so that when a user requests a route, the system can quickly retrieve and match pre-analyzed data rather than processing everything in real-time.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

Satellite images serve as an intermediary data source between the map data and the actual physical routes. The system extracts meaningful features (land cover, altitude, curvature) from satellite images to create a bridge between theoretical map routes and real-world conditions, enabling accurate route characterization without direct physical measurement.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If satellite image data is used to extract route characteristics, then the accuracy of route information is improved, but the data processing time and computational resources increase

Engineering Contradiction:

Improveaccuracy of environmental characteristicsVSAvoiddata processing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The satellite image processing is segmented into distinct feature extraction tasks: land cover classification, altitude analysis, curvature calculation, and environmental feature identification. Each segment can be processed independently and in parallel, reducing overall processing time while maintaining comprehensive analysis.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Route candidate areas are pre-identified and characterized from satellite images before user requests. The system performs preliminary extraction of shape characteristics and environmental features, storing them for quick retrieval and matching during recommendation, rather than processing images in real-time when users request routes.

Inventive Principle:

Principle #10Preliminary action

3Adaptability or versatility

If multiple route characteristics and environmental factors are analyzed, then the personalization and suitability of recommended routes are improved, but the complexity of analysis and clustering increases

Engineering Contradiction:

Improvepersonalization of route recommendationsVSAvoidcomplexity of clustering algorithm

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system replaces manual or simple rule-based route selection with automated machine learning clustering algorithms. The clustering model automatically learns patterns in route characteristics and environmental features, enabling sophisticated personalization and matching without requiring complex manual configuration or rule sets.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system transforms multiple diverse route characteristics (shape, environment, conditions) into standardized numerical parameters suitable for computational clustering. By converting qualitative features into quantifiable parameters, the system enables mathematical optimization and automated decision-making while maintaining adaptability to user preferences.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20230417563A1Method for providing recommended exercise route information, server and system using the same

Publication Date: 2023.12.28 AJOU UNIV IND ACADEMIC COOP FOUND

AI summary

Provided are a method of providing recommended exercise route information, the method comprises extracting exercise route candidate areas based on satellite image data collected, generating candidate exercise routes using the extracted exercise route candidate areas, analyzing shape characteristics and environmental characteristics of each of the candidate exercise routes, clustering, the candidate exercise routes based on the shape characteristics and the environmental characteristics of each of the candidate exercise routes, analyzing a condition of each of the candidate exercise routes using the satellite image data, providing recommended exercise route information to a user based on the clustered groups of the candidate exercise routes and the condition of each of the candidate exercise routes, wherein the environmental characteristics of the candidate exercise routes include architectural structures around the candidate exercise routes and natural environment around the candidate exercise routes, and the condition of each of the candidate exercise routes includes a state in which water is stagnant on the route due to rain, a state in which garbage exists on the route, and a state of construction on the route.