Development and Demonstration of AI-Based Light Pollution Image Analysis Solution

Overview

To prevent the harm caused by excessive artificial lighting at night, 17 metropolitan governments nationwide conduct light pollution environmental impact assessments every three years. However, as most of these assessments rely heavily on manual processes, they are unable to respond to persistent public complaints. Consequently, starting in 2024, support is being provided to an AI specialist consortium to develop and demonstrate AI solutions aimed at resolving issues in light pollution measurement methods, data analysis, and management.

Project description

To develop an AI solution, the project collected and processed target lighting data by applying light pollution measurement standards, thereby developing models for detecting target lighting in images, classifying lighting methods, and inferring luminance error ranges.

Additionally, the system added compatibility with area luminance measurement equipment and luminance data analysis capabilities, developed it as an AI solution, and validated it by linking it to a service platform.

Key services include automating the refinement and analysis of measured image data to detect target lighting through analysis of luminance and other characteristics. Furthermore, by establishing a light pollution environmental impact assessment database and linking data, it provides a service that derives relevant data and inference values during light pollution environmental impact assessments.

Key technological features

• Development of AI Base Models for: Advertising Light Detection/Classification and Luminance Error Inference
• AI solution development: Added compatibility with luminance measurement equipment and luminance data analysis capabilities Demonstration in conjunction with a service platform equipped with map layers

Expected outcomes

By applying AI solutions to measure and analyze the hazards of excessive lighting in areas lacking specialized personnel, this project is expected to reduce the time required to build an environmental impact assessment database. This will enable the establishment of a systematic light pollution management system, allowing for effective responses to light pollution-related public complaints. Furthermore, through future public service provision, the project will be able to build a favorable lighting environment.

Launch year: 2024

Attribution

Case study courtesy of the National Information Society Agency, Republic of Korea, Apolitical's Content Partner. Sourced from Use Cases of Public AI Service, Vol.1,  https://eng.nia.or.kr/site/nia_eng/ex/bbs/View.do?cbIdx=31975&bcIdx=28997&parentSeq=29046