Using High-Frequency Information and RH to Estimate AQI Based on SVR

Area of Science:

• Environmental Science
• Computer Vision
• Data Science

Background:

• Air pollution monitoring is crucial for public health.
• Existing methods for Air Quality Index (AQI) estimation can be complex and costly.
• Visibility is a key indicator affected by air quality.

Purpose of the Study:

• To develop a simple, fast, and low-cost method for estimating AQI.
• To explore the relationship between image-derived visibility and AQI.
• To create a predictive model for AQI using image analysis and environmental data.

Main Methods:

• Utilized images with varying PM2.5 concentrations to identify Regions of Interest (RoI).
• Calculated high-frequency information from RoI pixels.
• Employed Support Vector Regression (SVR) to train a model using RoI high-frequency information, relative humidity (RH), and true AQI.

Main Results:

• A significant positive correlation was observed between visibility and AQI.
• Decreased visibility was associated with increased AQI values.
• The proposed method demonstrated acceptable performance in estimating AQI.

Conclusions:

• Image-based visibility estimation offers a viable approach for AQI monitoring.
• The developed SVR model provides a simple, fast, and low-cost solution for AQI estimation.
• This method can contribute to more accessible environmental air quality assessments.